| Social Network | Social Capital
You are on page 1of 500

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.


Rosmawati Ab Raub
Authors: ros@tmrnd.com.my
TM Research & Development Sdn Bhd, FRAT Cluster, Selangor, Malaysia
Title: Social network analysis in analyzing potential collaborators
Since the tragic events of the September 11, 2001 attacks in the United States, social network analysis has
increasingly been used to study terrorist networks. Terrorism, together with such crimes as drug trafficking,
armed robberies and fraud is called organized crime. Crime or criminal activities usually work in group which
needs more than two people to communicate and to perform their secret plan. Traditionally, crime investigators
or intelligent officer used investigation tools that are widely available in the market to provide graphical
representation of the data network in order to facilitate and speed up their investigation. This paper proposes the
combination of social network analysis strength which focuses on ego-centric network method and important
criteria for potential collaborators in identifying potential collaborators in telecommunication data set. Effectively
combining multiple sources of data can lead intelligent officer to discover patterns in order to help them be more
proactive in their investigations.
Session: Mathematical Models
Keywords: egocentric networks
Accepted: Yes

1 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Einat Achituv-Du-nour
The Hebrew University of Jerusalem, Department of Sociology, Mount Scopus, Jerusalem 91905, Israel
Authors: Yuval Kalish
The Hebrew University of Jerusalem, Department of Sociology, Mount Scopus, Jerusalem 91905, Israel
Amalya Oliver
The Hebrew University of Jerusalem, Department of Sociology, Mount Scopus, Jerusalem 91905, Israel
Title: Maximizing Opportunities – Interorganizational Learning in a Biotechnology Consortium
Biotechnology related discoveries and products are known to be based on interorganizational learning processes
(Liebeskind et al. 1996, Oliver 2001, 2004, Powell et al. 1996, 1998, 2004). Moreover, much of the learning
exchanges are embedded in inter-institutional relations – between university and industry scientists (Oliver and
Liebeskind, 1998; Liebeskind and Oliver, 1998; Zucker and Darby, 1998). Since the scientific work in the
biotechnology industry is based on high level of reliance on basic science discoveries, there is a high level of
dependency between the two institutional actors – universities and industry, while much of the interorganizational
learning exchanges are inter-institutional. Our project focuses on a national funded five years long biotechnology
Abstract: consortium in Israel. The consortium included scientists from 11 firms and 16 university teams who had various
learning interactions and projects.

Science-based consortiums offer a pool of opportunities for learning, coupled with a need to hold information
from possible competitors. Our study questioned whether there is a distinct leading pattern of interorganizational
learning in the various co-existing networks. We use theories of closure and structural holes to hypothesize abou
the different structures that emerge in different learning networks. Analysis using Exponential Random Graph
models supports the predictions.
Session: Knowledge Networks
Keywords: access to knowledge , inter-organizational learning
Accepted: Yes

2 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Robert Ackland
The Australian National University, Research School of Social Sciences, Coombs Building, #9, Canberra, ACT
0200, Australia
Authors: Timothy Phillips
The Australian National University, Research School of Social Sciences, Coombs Building, #9, Canberra, ACT
0200, Australia
Amanda Spink
Queensland University of Technology Faculty of Information Technology Brisbane QLD 4000 Australia
Title: Adaptive Sampling of Hyperlink Networks: An Application to Research on Social Inclusion in Australia
Digital networks such as the World Wide Web (WWW) can enable socially-excluded groups to connect with other
groups and with the mainstream, thus promoting participation and giving them “voice”. However it is an empirical
question as to whether the voices of the socially excluded are being heard on the Web – who are they talking to,
and who is listening? In this paper we present new research into social inclusion in Australia using hyperlink data
from a large-scale crawl of the .au domain (comprising approximately 10 million web pages). We use adaptive
sampling (e.g. Thompson and Seber, 1996) to identify online networks of websites representing socially-excluded
groups (e.g. Aborigines, ethnic minorities, people with disabilities) and then analyse the structure and
composition of these networks to provide insights into the role of digital networks in promoting social inclusion.

Thompson, S.K and G. A. F. Seber. (1996). Adaptive Sampling. New York: J. Wiley & Sons.
Session: Sampling methods
Keywords: sampling , web , social inclusion , snowballing
Accepted: Yes

3 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Silke Adam
University of Hohenheim Institut of Communication Science / Media Politics (540E) Dept. of Social Sciences
Fruwirthstr. 47 70599 Stuttgart Germany
Title: Symbolic networks – a connection of content and network analysis for studying mass media debates
The goal of this paper is to show how public debates within the mass media can be analysed as networks, which
are characterised by the portrayed actors and their communicative interactions. Such networks are media
constructed and thus represent the symbolic dimension of political or societal disputes. To study such symbolic
networks it is proposed to systematically connect quantitative, relational content with empirical network analysis.
Such a combination allows to understand debates not only as treasure troves of isolated actors and frames, but
reveals how actors interact with each other showing not only actors’ visibility, but also the interaction structure
and coalition formation. Such a combination also allows comparing the voices of different speakers, also called
frame-sponsors, present in mass media debates. Linking these frame-sponsors to the frames they put on the
agenda reveals the roles different actors play in the competition for audience’s attention and support. How a
combination of content and network analysis can show the visibility, the interaction structure, the coalition
formation and actors’ roles within the mass media is illustrated by analysing the debate on Eastern enlargement
of the European Union in German and French newspapers.
Session: Collecting Network Data
Keywords: content analysis , discourse , framing , public space , symbolic networks , mass media
Accepted: Yes

4 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

jimi adams
University of Pennsylvania, Population Studies Center, 239 McNeil Building, 3718 Locust Walk, Philadelphia, PA
19104, USA
Title: Where's the Beef? Reframing Religious Organizations' Discordant HIV-Related Messages
Researchers studying intra-organizational information flow frequently address questions of efficiency and
structure of formal and informal networks. In an examination of within-denomination HIV-related messages in
rural Malawi, I find considerable differences in the content and presentation at the local-congregational level as
Abstract: compared to the national-denominational level. This initially appears to be evidence of poor efficiency, and lack o
internal agreement within denominational networks. However, by refocusing the question from what organizations
say to where messages are said (and constructed), a picture of message concordance arises that more directly
matches the relevant local-network patterns.
Session: Qualitative Network Research: Methodological and Theoretical Issues
Keywords: organizations , religion , information flow , hiv/aids
Accepted: Yes

5 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Larissa Adler-Lomnitz
Authors: larissa@servidor.unam.mx
National University of Mexico
Title: Political Culture and Social Networks a Comparative Study
The aim of this study is to test a theoretical model propose by the author for analyzing and understanding
Chilean and Mexican political cultures by means and examination of history and anthropological field research.
This model combines structural analysis based on social networks (horizontal and vertical) with the description of
the symbol system that provides them with feedback. It is the grammar of power and how it is express. We find
that Chilean society is horizontal structure in social classes within which are found informal social networks that
may for ideological reasons become formalized in political parties, composed in turn of different networks arising
out of friendship, within generational cohorts, and groups based on regional origin, family neighborhood, school
and other shared interests. At the same time in Mexico we find a power system composed of vertical networks
led by a strong leader which controls them and which has given rise in the twentieth century to a dominating party
(PRI) based on vertically organized system controlled at the top by the president of the republic.
Session: Political Networks
construction hierarchy , non.hierarchical networks , political networks , social network , political culture , chile
and mexico , leadership and networks , horizontal/vertical networks
Accepted: Yes

6 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Filip Agneessens
Authors: Ghent University, Department of Sociology, Gent, Belgium
Rafael Wittek
University of Groningen/ICS Department of Sociology, Groningen, The Netherlands
Title: Personal Characteristics, Interpersonal Relations and Conflict Management Strategies in Organizations
Organizational research has paid considerable attention to the way employees react to situations of conflict with
colleagues. However, the majority of existing research has assumed that the choice of response is primarily
dependent on the characteristics of the source of the strategy. In order to get a more complete idea about the
strategies used by actors, this paper uses a social network analysis approach to considers how the combination
of both direct and indirect strategies – which involve third parties, not directly connected to the conflict situation –
are combined to respond to a situation of conflict. Based on a cluster analysis of 8 relational conflict items
between employees of a company (bilateral arguing, bilateral negiotiation, complain to others, complain to
superior, vertical blame, retaliation, avoidance and resignation), 4 empirical types of conflict management
Abstract: strategies are distinguished: 1) a direct conflict approach focusing on negotiation, 2) a retaliation strategy
involving indirect control approaches and direct avoidance, 3) a strategy of allowing, focusing on resignation, and
4) a strategy of forcing that involves complaining to others and arguing to the source. Subsequently, a dyadic
independence p2-model is used to consider how the characteristics of the source, the target, as well as the
situational context help explain the use of a particular conflict management strategy. The results show that both
individual characteristics of the source and the target in a conflict situation (such as age, informal power, and
dependence in the group as a whole), the difference between the source and the target on these characteristics,
as well as the specific relation between both actors (on advice and trust) help explain the choice of conflict
management strategy.
Session: Intra-Organizational Networks
Keywords: conflict , organizations , p2 , advice , trust , power , dependence
Accepted: Yes

7 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Claudia Aguilar
Universidad Autónoma de Barcelona, Antropologia, Facultat de Lletres-Edifici B, Bellaterra, Barcelona 08193,
Authors: Spain
Pau Vives
Universitat de Barcelona Geografia Humana Facultat de Geografia i Història C/ Montalegre, 6-8. 08001
Barcelona, Spain
Title: Identification through personal networks: beyond ethnicist discourse in BiH
The research we are developing in Bosnia and Hercegovina arises from a dissatisfaction with the terms used in
anthropology and other social sciences when talking about how people organize themselves and organize others
in social groups.
In Bosnia and Hercegovina the official or hegemonic discourse divide the Bosnian people between Serbian,
Croatian and Bosnjak or Muslims and make them situate themselves almost in an obligatory way in one of these
three categories regenerated in a war context. But in a urban society like for example Sarajevo, these simple,
imposed and old identitary distinctions appear to be unclear and do not allow us to comprehend social
We believe that individual identification is relational. That means that it is a process in continuous re-elaboration
through social relationships. That is why we consider interesting the study of structure, composition and change
in personal networks because they can offer us much information about people’s everyday reality where the
collective identifications are continuously in negotiation.
Session: Networks and Identifications
Keywords: personal networks , identification , bosnia and hercegovina
Accepted: Yes

8 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Mito Akiyoshi
Authors: mito.akiyoshi@gmail.com
Senshu University, Humanities, 2-42-23-310 Akebono, Tachikawa, Tokyo 1900012, Japan
Title: Family or Friends? : Sources of Support and Companionship in the U.S. and Japan
The present study attempts to contribute to a better understanding of the concept and reality of social capital by
exploring the prevalent types of social relations and support systems in two different societies, namely, the
United States and Japan. In the 2000s the concept of social capital, as most notably investigated by Robert
Putnam in his analysis of American community is getting incorporated into the vocabulary of Japanese policy
community. According to Putnam, social capital provides a solution to problems that fall under the rubric of
“collective action problems,” “the prisoner’s dilemma,” “the free-rider problem,” and “the tragedy of the
commons.” However, little is known about the cross-cultural similarities and differences in social capital creation
Abstract: and maintenance. Using cross-cultural survey data, the present study found that the two societies differ in the
make-up of dominant types of social relations. Some of the key findings are as follows. First, coresidence with
adult family members other than one’s spouse or partner is more frequent in Japan than in the United States.
Second, the Japanese are more likely to turn to their relatives both for companionship and for social support.
Finally, beyond family networks, memberships in various types of associations are explained by both ascriptive
and achieved characteristics in the United States whereas ascriptive characteristics are more dominant in Japan.
By way of conclusion, these findings suggest that social relations and support networks among Americans are
more elective than those among the Japanese, which are characterized by the importance of ascriptive identities
Session: Social Capital
Keywords: social support , social capital , family networks , friendship networks , social relationships
Accepted: Yes

9 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Angelos Alexopoulos
Authors: angelos.alexopoulos@dcu.ie
Dublin City University Business School Dublin Ireland
Title: An Integrative Approach to the Role of Social Capital in Interpersonal Knowledge Transfer
Recent research highlights the prominence of social relationships for energising knowledge flows within work
organisations. Yet, there is a lack of understanding of how the structural, relational and cognitive facets of social
capital (Nahapiet & Ghoshal, 1998) inherent in those relationships interact with each other in an information and
advice seeking context. This study constructs and tests a model for investigating the configurational impact of
individual-level social capital on effective knowledge transfer. Using standard egocentric network techniques,
data on 270 knowledge exchange dyads were collected from employees within divisions of three Irish
knowledge-intensive organisations. The results of hierarchical multiple regression provide strong support for the
Abstract: proposed model. First, it was found that the positive effect of tie strength on knowledge transfer was mediated by
cognitive social capital as this is expressed in shared language and common values and goals. Second, the
effect of cognitive social capital on knowledge transfer was, in turn, mediated by relational social capital
expressed in reliance-based trust and disclosure-based trust. Third, while both types of trust appeared central to
the transfer of mainly codified knowledge, disclosure-trust emerged particularly important for the effective transfer
of non-codified knowledge. The study advances prior research (e.g. Levin & Cross, 2004) by offering a
multidimensional view of the interconnections between distinct features of social capital utilised for effective
transfer of knowledge at the dyadic level of analysis.
Session: Social Capital
Keywords: social capital , knowledge transfer , egocentric network , knowledge exchange dyads , trust
Accepted: Yes

10 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sara Nuzhat Amin

Authors: McGill University, Sociology, Apt 103, 410 Sherbrooke Est, Montreal, Quebec H2L 1J6, Canada
Tanya Trussler
McGill University, Sociology, Apt 103, 410 Sherbrooke Est, Montreal, Quebec H2L 1J6, Canada
A positional analysis of violent Islamic political extremism: mapping the connective structures of terrorist
Recent political events have increased interest in the way in which terrorist groups are structured and terrorists
are created. Using the MPIT Terrorist Knowledge Database and Michael Sageman’s ethnographic material on
Al-Quaeda terrorist networks, this paper employs a network perspective to examine two facets of violent political
extremism among Islamic groups from 1972-2005. First, we map the connective structure of all leaders and
members of Islamic organizations who have been identified as linked to terrorist activity over the last 33 years.
Second, we ask how do these connective structures or networks change over time, specifically, whether terrorist
networks remain diffuse/decentralized or do they tend to become hierarchical? Finally, we use a positional
analysis to examine the significance of centrality for individuals to commit acts of terrorism vs. play leadership
and organizational roles. For instance, we examine whether those who are most connected are the ones most
likely to commit acts of terror, or those who lie at the periphery of the Islamic terrorist network.
Session: Criminals, Gangs, Terrorists, and Networks
affiliation networks , associational membership , differential association , embeddedness , horizontal/vertical
networks , hierarchy , network change , terrorist networks
Accepted: Yes

11 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

George Anghelcev
University of Minnesota, School of Journalism and Mass Communication, 330 Murphy Hall, 206 Church At. SE,
Minneapolis, MN 55455, USA
Itai Himelboim
University of Minnesota, School of Journalism and Mass Communication, 330 Murphy Hall, 206 Church At. SE,
Minneapolis, MN 55455, USA
Julie Jones
University of Minnesota, School of Journalism and Mass Communication, 330 Murphy Hall, 206 Church At. SE,
Minneapolis, MN 55455, USA
Network Analysis as a Marketing Tool: Using Network Parameters to Identify. Brand Zealots in Online Consumer
In recent years, interpersonal communication networks have begun to replace advertising as the primary channel
of persuasion about products and brands. This declining interest in advertising is generally attributed to
increased clutter, reduced consumption of traditional media and spread of technologies (such as TiVO) which
enable consumers to avoid exposure to ads. Consequently, businesses have begun to use interpersonal
networks of influence for disseminating information about their products and brands. To achieve best results,
marketers need to identify market opinion leaders, or marketing leaders. Brand Zealots (currently 4% of total US
consumers) are specialized opinion leaders for whom opinion leading activities related to their favorite brands are
center-stage in interpersonal communication (Eighmey et al, 2005). They engage in advocacy on behalf of their
favorite brands and, once identified, can amplify a brand’s message by disseminating it in the social networks in
which they are engaged.
The strategic use of marketing leaders is hindered by the high cost and duration needed to reach and identify
them in the marketplace. The availability of online communication logs and networks, allows marketers to
overcome these obstacles.

This paper suggests a conceptual and methodological framework, to identify marketing leaders, namely brand
zealots, in online brand communities such as discussion groups, forums and blogs. These online platforms take
the form of networks, via interpersonal interactions and exchange of hyperlinks. We advance that, based on
measures of centrality and prestige, coupled with computerized content analysis, one can identify brand zealots
in online environments, at low costs, in order to further use them as central pieces of marketing promotional
Session: Networks, Economics, and Markets
Keywords: internet , online discussion boards , advice
Accepted: Yes

12 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Gilda Antonelli
Authors: University of Molise, Department of Management, Via De Sanctis, Polifunzionale II, Campobasso 86100, Italy
Raffaele Corrado
University of Bologna, Department of Management, Via Capo di Lucca 34, Bologna 40126, Italy
Title: Strategies for Legitimating a Broker. The Interplay of Social Capital and Organizational Role.
This work is about the co-evolution of intra-organizational networks and organizational structures, in the context
of the introduction of an innovation. Network studies concerned with social capital in organizational settings
focused on the brokerage role of actors that connect otherwise disconnected sets of network actors. These
studies document the consequences of the broker network position, but do not analyze the mechanisms through
which this position is acquired vis a vis the other actors in the network. Our aim is assessing the mechanisms
through which the broker legitimizes himself in his own role.

Our hypotheses can be summarized as follows: an actor acquires a brokerage role as a consequence of his
informal network position and relational capabilities, partially mediated by his formal role in the organization. We
analyze the dynamics of a network created for the implementation of a technological innovation with important
organizational implications. Our analysis combines ethnographic participant observation with the formal analysis
of network evolution, based on Tom Snijders’ SIENA models.

The analyzed networks include about 40 individuals and 4 types of tie, observed through sociometric
questionnaires at 3 time points that correspond to distinct phases of the implementation of the innovation.
Questionnaires were complemented by direct observation and interviewing of the actors. The empirical context of
this research is a small Italian university with about 10,000 students, 2,550 employees and approximately 500
teaching staff members.
Session: Intra-Organizational Networks
Keywords: intraorganizational networks , functioning of social capital , innovation
Accepted: Yes

13 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Belinda Aparicio Diaz

Vienna Institute of Demography, Austrian Academy of Sciences, Vienna, Austria
Thomas Fent
Authors: Vienna Institute of Demography, Austrian Academy of Sciences, Vienna, Austria
Alexia Prskawetz
Vienna Institute of Demography, Austrian Academy of Sciences, Vienna, Austria
Laura Bernardi
Max Planck Institute for Demographic Research, Rostock, Germany
Title: Transition to parenthood: The role of social interactions and endogenous networks
This paper investigates how the decision of having an additional child is being influenced by an individuals peer
group. We show via agent based simulations how social interaction creates interdependencies in the individual
transition to parenthood and its timing. We build a one-sex model, and provide agents with four different
characteristics. Based on theses characteristics agents endogenously form their network. Network members then
may influence the agents' transition to higher parity levels. The agents compare the share of agents with a higher
parity than their own within their peer group with the same share on the aggregate level. Our numerical
simulations indicate that accounting for social interactions is important to explain the shift of first birth
probabilities in Austria over the period 1984 to 1994.
Session: Social Networks and Behavioral Change
agent based modelling , fertility , social learning , social networks , peer networks , perceived social
networks , kinship network , transition to partenthood
Accepted: Undetermined

14 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sinan Aral
Authors: NYU / MIT, Information Systems, 44 West 4th Street Room 8-81, New York, NY 10012, USA
Marshall Van Alstyne
BU / MIT Information Systems, Boston, MA, USA
Network Structure & Information Advantage: Structural Determinants of Access to Novel Information and their
Performance Implications
We examine relationships between social network structure, information structure, and individual performance.
Specifically, we investigate which network structures influence access to diverse and novel information, and
whether these relationships explain performance in information intensive work. We build and validate an
analytical model of information diversity, develop hypotheses linking two key aspects of network structure - size
and diversity - to the distribution of novel information among actors, and test our theory using empirical evidence
from a ten month panel of email communication patterns, message content and performance among employees
of a medium sized executive recruiting firm. Our results indicate that: (1) the total amount of novel information
and the diversity of information flowing to actors are increasing in their network size and network diversity.
However, (2) the marginal increase in information diversity is decreasing in actors’ network size, a result
explained in part because (3) network diversity is increasing in network size, but with diminishing marginal
returns. (4) Network diversity contributes to performance even when controlling for the positive performance
effects of access to novel information, suggesting additional benefits to network diversity beyond those conferred
through information advantage. Surprisingly, (5) traditional demographic and human capital variables have little
effect on access to diverse information, highlighting the importance of network structure for information
advantage. The methods and tools developed are replicable and can be readily applied to other settings in which
email is widely used and available, opening a new frontier for the analysis of networks and information content.
Session: Knowledge Networks
content analysis , information benefits , information content , information diversity , knowledge transfer ,
structural holes , productivity , information work
Accepted: Yes

15 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Rodrigo Araya
Authors: raraya@gmail.com
Research group Egoredes - Universidad Autónoma de Barcelona www.egoredes.net
Multitudes and networks: approach to the study of collective action from the perspective of social networks.
Belgrade 2000 case study
The paper focuses on the role of personal networks in demonstrations against the repressive regime of Slobodan
Milosevic in October 2000.

The central aim of the paper is to analyse past demonstrations and new patterns of collective action from the
Social Networks perspective. The basic assumption is that collective action can be classified in three ideal types
of networks: centralized, decentralized and distributed. This paper proposes that the new patterns of collective
action are related with distributed networks, which are recent phenomena in society mainly because of
communication technologies development.

Distributed networks are a special configuration of ties where theoretically any node can be connected with any
other, without filters, intermediation or manipulation.

This paper proposes that distributed networks are an emergent pattern in collective action which becomes
possible only in critical situations, such as natural catastrophes, extreme political crisis, among others.

In day to day life social networks are conditioned by several social rules. In critical situations people feel free to
use all communication channels to spread information and recruit wills for the collective action.

In terms of personal networks (micro level), we have observed two situations. Firstly, people put their personal
networks (as a resource) for the benefit of collective objectives, e.g. to help victims in natural catastrophe, to fight
against a dictatorship. Secondly, people might use latent contacts (acquaintances, weak ties, sleepy
relationships) which in normal situations are inactive. In brief, there is more disposition to extend the personal
network through new ties (temporarily or permanently) or activating latent contacts.

In Belgrade demonstrations, like in East Germany’s October 1989 (Opp and Gern, 1993), personal networks
were the most important contexts for mobilizing citizens, more than mass media, political parties or any other
formal organization.

The empirical approach considers two main research techniques: ethnographic observation and personal
networks. The ethnographic observation has been combined with short life stories (centred in the nineties). On
the other hand, 50 personal networks of people involved in demonstrations have been analyzed. This research is
still in progress.
Session: Qualitative Network Studies I
Keywords: collective action , personal networks
Accepted: Yes

16 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Charles Armstrong
Authors: charles@trampolinesystems.com
Trampoline Systems Ltd, London, UK
Title: Harnessing Social Networks for Emergent Team Formation in the Enterprise
Traditional communities are extremely efficient at clustering relevant people around discussions, decisions and
tasks of importance to the collective. These processes operate with negligible formal structure and no top-down

In contrast large business organisations require explicit management inputs to signal the need for a team,
identify the people who should be on it then monitor its progress. This formal mechanism appears to be
significantly less efficient at engaging the available intelligence and expertise than the informal process found in a
traditional community.

In this session Charles Armstrong examines how social networks enable efficient cluster-formation in a traditiona
community. He goes on to outline how technologies such as Trampoline's SONAR platform maps collaboration
networks in the enterprise by mining data from installed information systems. Finally Charles proposes how this
intelligence can be harnessed to enable emergent team formation in the enterprise and assign resources more
Session: Business Networks
Keywords: emergent structure , data mining
Accepted: Yes

17 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Claus Atzenbeck
Authors: Aalborg University Esbjerg, Department of Software and Media Technology, Esbjerg, DENMARK
Petra Deger
University of Regensburg, Institute of Sociology, Regensburg, GERMANY
Title: Representation of social networks from an actor's point of view
Analyses of social networks usually focus on objective representations. However, mostly they are based on a
simple graph-based model which only represents connections between people, but ignores most or even all of
their contexts, such as the kind of relationship. We aim to extend this model by investigating how actors perceive
and represent their own networks. From an ethnomethodological point of view, we survey social order from an
actor's perspective. This is called egological network. Qualitative experiments and interviews provide data which
allows us to draw conclusions about people's perception of such networks. They also let us discover which
information actors ignore or use for representing their own social connections. Our findings so far let us assume
that the actor's point of view on social networks is more complex than simple links could represent conveniently.

We understand our approach as an analogy to developments in computer science, especially hypertext. Simply
Abstract: spoken, hypertexts are informational units (nodes) connected through links. However, these relations are explicit
and strict, similar to graph-based social networks. In order to provide a less explicit way of structuring,
researchers developed spatial hypertext. Following a cards-on-table metaphor, associations are represented by
visual cues, distance, or alignment of information snippets on a space. This applies to user's visual recognition
and supports the emergence of structure over time. Spatial structures allow humans to express and understand
complex multilayered facts and subtle differences often more easily than with node--link structures or generic

We aim to improve egological social networks by applying findings from hypertext and structure domain research
We approach this goal in an interdisciplinary fashion and expect helpful synergistic effects from our collaboration
of social and computer scientists.
Session: Visualization
actor agency , ethnomethodology , egocentric network , experiment , hypertext , spatial structure ,
visualization , interdisciplinary approach
Accepted: Yes

18 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Anna-Liisa Aunio
Authors: annaliisa.aunio@gmail.com
McGill University Department of Sociology Montreal, Quebec, Canada
Title: Climate Action Networks for Change: Mobilizing Coalitions and Networks in National and International Politics
This study explores the relationship between networks and coalitions in national and international venues through
an empirical examination of the Climate Action Network (CAN) in Canada and the United States. CAN is an
international sustained coalition composed of environmental nongovernmental organizations (ENGOs) that
campaign to reduce human-induced climate change in local, national, regional, and international contexts. CAN
members have a mandate to coordinate their activities on climate change through the coalition, but most
members do so only occasionally. I ask: when do members of CAN work through CAN and, alternatively, when
do they work through other networks in their campaigns on the issue of climate change? Through an analysis of
affiliation networks on climate change policy in both Canada and the U.S. from 2001-2005, I test several
hypotheses that may account for variation in network activity, including location, issue orientation, and network
centrality. By explaining variation in network activity through CAN, social movement and network scholars can
gain a purchase on what distinguishes a sustained coalition from advocacy network, the circumstances in which
they become active, and what consequences it has for collective action.
Session: Policy Networks on Climate Change
Keywords: coalitions , activism , environmental networks
Accepted: Yes

19 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Christine Avenarius
Authors: avenariusc@ecu.edu
East Carolina University, Department of Anthropology, 1401 E. 5th Street, Greenville, NC 27858, USA
Title: To Bribe Or Not To Bribe: Social Networks and Attitudes Towards Corruption In Rural China
Corruption is one of the most critical issues facing contemporary China, influencing not only the distribution of
wealth and opportunity but also the legitimacy of government. Contrary to expectations economic liberalization
starting in the early 1980s has not reduced corruption but rather amplified its frequency and scope. Despite
efforts by the Chinese government to establish the rule of law and construct a new legal system, the rule of
relationships continues to influence the daily reality of Chinese citizens. Although Chinese people today can take
their disputes to a formal mediator or a judge at court, few take advantage of this opportunity. In fact, the analysis
of narratives about perceptions of justice and fairness collected from urban and rural residents in Hebei province
Abstract: revealed that most informants believe that to win a case at court they have to channel gifts to the presiding judge
The inevitability of gift giving to accomplish goals divides citizens into two primary types of social actors: those
who refrain from further involvement and those willing to make the necessary payments. Money is not all that
matters, however. Structural positions in personal and whole community networks effect the decision to risk
going to court. This paper combines findings from a social network study of 184 households in Li Village with 56
in-depth interviews with heads of households regarding their attitudes towards justice and corruption. It
demonstrates that relationships to people outside a rural community are the main predictor of willingness to get
involved in corrupt practices.
Session: Social Support
actor agency , civil society , community structure , conflict , data collection , farming community , inequality ,
social capital
Accepted: Yes

20 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Javier Augusto Avila Molero

Ph.D Candidate, Group Egoredes, Universidad Autonoma de Barcelona, Barcelona, Spain.
Title: Personal networks and acculturation: The case of Argentineans in Barcelona
The international migration process implies a radical change in structure and composition of immigrant's persona
networks. Once the immigrants arrive to the new society they have to adapt their personal networks to the new
situation, frequently rebuilding both a new core of strong ties and a new periphery of weak ties. Our first
hypothesis is that this process results in one of four types of personal networks: a) transnational, b) segregated,
c) assimilated and d) cosmopolitan. Each one of these has a particular structure and composition. The second
hypothesis is that there is a relation between the level of heterogeneity of the immigrant personal network and
the kind of process of acculturation developed. This paper is based on 40 structured interviews with Argentinean
migrants in Barcelona collected in 2006 using egonet and its visualization capabilities.
Session: Personal Network Methods
Keywords: personal networks , immigrants , acculturation
Accepted: Yes

21 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Istvan Back
Authors: University of Groningen, Department of Sociology/ICS, Grote Rozenstraat 31, Groningen 9712 TG, Netherlands
Andreas Flache
University of Groningen, Department of Sociology/ICS, Grote Rozenstraat 31, Groningen 9712 TG, Netherlands
Title: Commitment and Networking under Uncertainty
A wealth of anthropological evidence suggests that people protect themselves against future risks and
uncertainty by increasing their network of partners who can be called upon in times of need. On the other hand,
comparison of various markets reveal a negative association between the number of partners and uncertainty, as
do laboratory experiments with controlled exchange. In the current work we propose an explanation for this
Abstract: paradox by separating different types of uncertainties. In particular, while social uncertainty concerns the
trustworthiness of a potential partner, resource uncertainty concerns the partner’s resources. We argue that
although both social and resource uncertainty have the potential to increase commitment on their own, they
interact and can even cancel out each other's effect. We present laboratory experiments, as well as on-line
experiments that give empirical support to our hypotheses.
Session: Networks and Trust
egocentric networks , exchange networks , experiments , game theory , network strategy , relationship
formation , trust , uncertainty
Accepted: Yes

22 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Wendi Backler
The Boston Consulting Group, Strategy Practice, 5343 Cliffridge Avenue, North Vancouver, BC V7R 3V5,
Network Competitive Advantage: How A Network Perspective Can Point the Way To Winning Strategies in
Business performance today depends on increasingly complex and often informal networks of interactions and
relationships. These networks exist within the company (e.g. knowledge networks of “go-to people”) and extend
well beyond its legal boundaries (e.g. supply chain participants, customers, influencers, regulators, etc.). Much
work has been done to understand and ‘improve’ existing networks within and between organizations and within
and between the people that comprise them, resulting in stronger organizations, and more and better
collaboration. But evidence is starting to emerge that the real power of networks in a business context comes
from how they position a company to ‘win’ against its competitors.

This paper presents a preliminary conceptual framework about how networks can convey advantage from a
business strategy perspective. Evidence is emerging from a variety of perspectives, including academic
research, case studies and BCG’s own experience in the ‘client lab,’ to support the notion that companies can
gain sustainable competitive advantage from networks in three main ways:
-- Positional advantage - Organizations that ‘bridge’ multiple networks have been shown to have both a
performance advantage and creativity and learning advantage. ‘Hubs’ have the ability to exercise power and
influence over the network, and control or accelerate flows of information.
-- Structural advantage - Patterns of connections impact the efficiency, effectiveness and acuity of information
flows within and between organizations. It is well documented that certain network topologies are better at some
things than others. Businesses in any given industry are actually ‘networks competing against networks’, and
therefore a network perspective can help both explain and predict advantage.
-- Adaptability advantage - How links form and reform can optimize organizational flexibility to respond to a crisis
or a rapidly changing external environment. The dynamics of interactions between individuals, teams and
departments can vary dramatically between organizations, as can the strength and nature of ties.

Examples of each of the above are provided in the context of innovation networks and influence networks.
Session: Business Networks
business networks , network competitive advantage , power in networks , influence , positional advantage ,
structural advantage , adaptability advantage
Accepted: Yes

23 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Chris Baerveldt
Utrecht University, Pedagogiek, PO Box 80.140, Utrecht 3508 TC, Netherlands
Bonne Zijlstra
Authors: University of Amsterdam Amsterdam, The Netherlands
Muriël De Wolf
Ghent University Ghent, Belgium
Ronan Van Rossem Marijtje van Duijn
Ronan: Ghent University, Ghent, Belgium Marijtje: Universty of Groningen,The Netherlands
Title: Ethnic boundaries in Belgian and Dutch student’s networks.
Ethnic boundaries were tested in students’ networks in 34 Belgian and 19 Dutch high schools. Each network
consisted of a school cohort in an intermediate level of education (track). While students from the native majority
predominantly had friendships within their own ethnic category, minority students often had more inter-ethnic
than intra-ethnic friendships. However, a multilevel p2 model for analyzing the networks showed that this was
caused mainly by the quantitative dominance of native students in the networks. Native students were much less
inclined to choose for inter-ethnic friendships than minority students. We found ethnic boundaries to be stronger
in the Dutch networks than in the Belgian networks. Although this may be partly due to methodological reasons, it
is still surprising that the boundaries in the Belgian networks were not stronger. The Dutch data stem from the
pre-9/11 days when the Netherlands were still known for its tolerant climate; while the Belgian data stem from
2005 when one out of every 4-5 Belgians voted for an ethnocentric party.
Session: Networks and Identifications
Keywords: ethnic identity , adolescence , social networks , segregation , friendship networks
Accepted: Yes

24 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Delia Baldassarri
Authors: db2237@columbia.edu
Columbia University, Sociology, 435 West 119th Street apt. 9J, New York, NY 10027, USA
Title: Associative Patterns and Political Polarization in the US (1972-2004)
I relate trends in associational participation to dynamics of political polarization. I use GSS (1972-2004) data on
organizational involvement to focus on individuals' patterns of multiple affiliations across different type of
associations. I relate associational types on the basis of their shared members and trace the evolution of the
inter-organizational network over time. Similarly, I trace the level of intra-organizational and inter-organizational
heterogeneity on the basis of the socio-economic characteristics and political attitudes of their members.
The analysis confirms the decline in associational life, but it also suggests that the decline in civic engagement
Abstract: may not have had any consequence on the overall level of social integration. In general, if patterns of affiliation
show increasing socio-economic homophily and ideological alignment, then we would conclude that Americans
are experiencing increasingly overlapping social spheres, and thus that the civil society as a whole is on a
divisive path. In contrast, if there are instances of growing associational heterogeneity and ideological
de-alignment, than we would conclude that Americans are embedded in increasingly cross-cutting social
spheres, fostering a more integrated civil society. Preliminary results suggest that the civil society is not
becoming increasingly divided, although there is not sufficient support (yet) for the latter hypothesis.
Session: N/A
Keywords: affiliation networks , civil society
Accepted: Yes

25 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Artur Baldauf
Authors: University of Bern, Dep. of Management, Engehaldenstrasse 4, Bern, Bern 3012, Switzerland
Olaf Rank
University of Bern, Dep. of Management, Engehaldenstrasse 4, Bern, Bern 3012, Switzerland
Title: Organizational Antecedents and Performance Consequences for Network Embeddedness in Sales
The study analyzes the effects of managerial supervising activities on the embeddedness of salespeople into
cooperative relationships. Specifically, behavior and outcome control as well as perceived organizational support
are examined as antecedents of salespeople’s willingness to exchange business-related information and
know-how. Moreover, the effects of network embeddedness on the performance of salespeople are studied. Data
has been gathered from 288 salespeople working for nine sales units in five companies. The companies had
different industry backgrounds (electronic parts and components, energy supply, postal services, banking, and
insurance) and were based in Germany and Switzerland. The size of the sales units varied substantially ranging
Abstract: from 13 employees for the smallest unit to 81 in the largest. The results reveal that behavior control is negatively
related to network embeddedness whereas positive effects can be stated for outcome control and perceived
organizational support. Moreover, significant interaction effects can be found between the two forms of
managerial control and perceived organizational support. Finally, it can be shown that higher levels of network
embeddedness are clearly associated with increased job performance. A key implication of the study is that
salespeople’s embeddedness into collaborative networks is of central importance in the context of supervision
and performance. Network embeddedness fully mediates the effects of behavior and outcome control on job
performance and partially mediates the relationship between perceived organizational support and performance.
Session: Intra-Organizational Networks
Keywords: business networks , effectiveness , embeddedness , intraorganizational networks
Accepted: Yes

26 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Prasad Balkundi
SUNY Buffalo, Organization and Human Resources, 274 Jacobs Mgt Ctr, Buffalo, NY 14260, USA
Judd Michael
Sustainable Wood-based Enterprises 211 New Forest Resources Building Penn State University University Park,
PA 16802-4703 USA
Zoe Barsness
Milgard School of Business University of Washington, Tacoma 1900 Commerce Street USA Tacoma, WA 98402
Title: The leadership Paradox: How leadership network structures predict turnover intentions among team members
How does leadership change group dynamics and individual team
members’ commitment to their teams? This study seeks to answer this
question by testing for the association between leadership network
structures, team conflict and intentions to quit the team. Based on the
data collected from 236 employees working in 20 teams in a manufacturing
organization we tested for cross-level mediation using Random
Coefficient Modeling. Teams with central formal leaders (measured in
terms of in-degree centrality in the advice network) experienced lower
levels of team conflict and members had lower intentions of quitting
their team. In contrast, betweeness centralization among subordinates
(i.e., the presence of informal leaders among subordinates) was
associated with elevated levels of team conflict and higher turnover
intentions. Team conflict partially mediated the effects of leadership
network structure on turnover intentions.
Session: Intra-Organizational Networks
Keywords: leadership and networks , centrality , informal networks , teams , turnover
Accepted: Yes

27 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Paul Bance
University of Nottingham - School of Politics - Centre for the Study of Global and Social Justice - Nottingham,
Title: Social Movements and Virtual Networks: Researching the Possible Contribution of Hyperlink Network Analysis
Social movement scholarship has increasingly emphasized the network nature of its object of research in recent
years. Once recognized the significance of the concept of network in the study of social movements, a major
methodological issue faced by researchers is the systematic collection and coding of data that will permit
subsequent analyses. This paper examines the hypothesis that established methods of investigation may be
complemented by the use of new techniques that focus on virtual networks to understand the structure and
dynamics of movements. The rationale for this argument rests on the following idea: any component of a social
movement has also a digital existence today (e-mail, web site, etc); then, if there is a social movement, and thus
a network, these virtual realities must be interconnected; therefore, exploring the cyberspace should permit to
produce an additional picture of the movement. Computer-assisted collection, coding and visualization of data
relative to the hyperlink structure of two case studies – the European movement against migrant discrimination
and the UK anarchist movement – are performed to test the validity and reliability of this research approach, i.e.,
virtual network analysis applied to social movement studies. This method shows good results for the two
movements under consideration. Visual illustrations (hyperlink maps) are included to conclude and discuss the
promises and drawbacks of such an approach.
Session: Collective Actions and Social Movements
Keywords: activism , data collection , internet/hyperlink networks , social movements , visualization , virtual network
Accepted: Yes

28 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Emma Barnes
Authors: emma.barnes@nuim.ie
Department of Sociology, National Univeristy of Ireland Maynooth, Co Kildare, Ireland
Title: The two fold nature of social supports
The focus of this paper is the two fold nature of the sources of support and the impact on mental well-being.
Social supports can have both positive and negative consequences. The paper will explore the experience and
the impact of these relationships on one's well-being and mental health. The analysis is drawn from a selection of
Abstract: in-depth interviews and network analysis from the first round of interviews with young adults aged 18 to 30 years.
The study consists of two waves of interviews which are six months apart. The research draws upon two groups
of young adults. Those with mild depression who have sought treatment and those with no mental health
Session: Social Support
Keywords: depression , negative ties , social support , positive ties
Accepted: Yes

29 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Roy Barnes
Authors: rcbarnes@umflint.edu
University of Michigan-Flint, Department of Sociology, 303 East Kearsley Street, Flint, MI 48502-1952, USA
Title: The Organizational Mosaic of U.S. Corporate Directors
This paper describes the changing network structures among organizations that are formed by interlocking
directors of major corporations in the United States between 1962 and 1995. The analyses begin by describing
the networks based on corporate ties alone. However, because corporate directors are also members of such
organizations as non-profit foundations, cultural organizations, university boards of trustees and social clubs, the
research explores how such additional social affiliations alter the network characteristics among these
organizations over the second half of the Twentieth Century.
Session: Politics and Interlocking Directorships
Keywords: inter-organizational , interlocking directors
Accepted: Yes

30 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

George Barnett
Authors: gbarnett@buffalo.edu
SUNY@Buffalo, Communication, Department of Communication, SUNY @Buffalo, Buffalo, NY 14260, USA
Title: An Examination of the Changes in the International Telecommunications Network after September 11, 2001
This paper describes changing pattern in the international telephone network using data from the International
Telecommunication Union’s World Telecommunication Indicators Database. It describes how certain countries’
incoming and outgoing international telephone traffic (in-degree and out-degree centrality) changed since 1980
paying particular attention to the period around 2001. The results indicate that there was a alteration in the
ongoing rate of change in the international telecommunication network around the time of the September 11,
2001. While the terrorist attack and reactions to those events offers one explanation for these changes,
alternatives are suggested that involve the diffusion of international telephony and voice communication over the
Internet (VoIP).
Session: International Networks
Keywords: globalization , world systems
Accepted: Yes

31 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Vladimir Batagelj
Authors: vladimir.batagelj@fmf.uni-lj.si
University of Ljubljana, FMF, Matematika, Jadranska 19, Ljubljana 1000, Slovenia
Title: Viszards Session
Viszards sessions started at Sunbelt XXII with
different analyses and visualizations of the media
coverage of the September 11 terrorist attacks.
At the following Sunbelts we continued with:
XXIII: 'The Summer Joker' network; XXIV: the
players market of the football World Championship
2002; XXV: KEDS (The Kansas Event Data System);
and XXVI: IMDB (The Internet Movie Database)

With this year's session we are continuing this

tradition. Rather than a series of contributions on
related subjects, this session features a single,
joint presentation by all contributors. Our aim is
to demonstrate the richness and power of network
analysis, in particular when supported by
visualization. We therefore present a
multi-perspective analysis of a single data set,
utilizing a broad range of visualization methods.

This year we will analyze Wikipedia networks -

http://www.wikipedia.org/ .
Session: Viszards session
Keywords: methodology , network structure , visualization , citation network , wikipedia
Accepted: Yes

32 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Vladimir Batagelj
University of Ljubljana, FMF, Matematika, Jadranska 19, Ljubljana 1000, Slovenia
Authors: Patrick Doreian
University of Pittsburgh Department of sociology Pittsburgh, USA
Anuška Ferligoj
University of Ljubljana Faculty of Social Sciences, Kardeljeva pl. 5 Ljubljana 1000, Slovenia
Title: Three Dimensional Blockmodeling
This paper presents an effort to apply blockmodeling to
three dimensional network structures. Such data arise
naturally in many situations. The approaches to blockmodeling
such structures that we propose and develop are fourfold:
indirect methods, coupled direct and indirect methods,
graph theoretical methods, and full generalized blockmodeling.

A formalization of the three dimensional blockmodeling problem

Abstract: is presented together with a formal statement of the methods
for solving this problem. The software support for indirect
(structural equivalence) and direct (generalized blockmodeling)
approach was developed.

These methods are applied to a variety of real three dimensional

network data sets. Three dimensional displays (in kinimage
format) of the obtained partitions are used to present
the results.
Session: Blockmodeling and clustering of networks
blockmodeling , clustering , network structure , social network analysis tools , visualization , three dimensional
networks , algorithm
Accepted: Yes

33 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dominik Batorski
Authors: Warsaw University, Insitute of Sociology, Karowa 18, Warsaw 00-324, Poland
Pawel Kucharski
Warsaw University, Insitute of Sociology, Warsaw, Poland
Title: The diffusion of innovations in large social networks
Most studies regarding complex social networks concentrate on the structural properties and the evolution of the
network itself. Relatively little research has been done on the social processes that are taking place through
interpersonal relations and communication. Yet processes like the spread of information, ideas and norms are
crucial within societal dynamics. This research is dedicated to explaining how one such process, namely the
diffusion of innovations, occurs in large-scale social networks. Often a contagious process, here, one’s decision
to adopt the given innovation depends on the behavior and opinions of the immediate others.

Within this research we investigate how different global properties of networks (i.e. short paths, scale free degree
distribution) can affect the diffusion process. We also focus on the mechanism responsible for adoption of the
innovation on the local level – personal network properties and personal network exposure (i.e. the degree an
individual is exposed to an innovation through his/her personal network). Additionally we compare relational
(cohesion) and structural (structural equivalence) network diffusion models.

In our analysis we use the networks of relations among almost 6 million users of Gadu-Gadu and almost 1 million
users of Grono.net – the most popular instant messenger and social networking website in Poland. Both Grono
and Gadu-Gadu periodically implement new features such as voice calling into their own services, which we
categorize as innovations, and then explain the diffusion of their usage. The anonymized whole network data
were provided by the owners of Gadu-Gadu and Grono.
Session: Network Dynamics
contagion , diffusion , dynamics on networks , innovation , internet , network processes , social influence ,
small world
Accepted: Yes

34 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Martina Battisti
Massey University Department of Human Resource Management 11222 Private Bag Palmerston North, New
Authors: Zealand
Christof Netzer
Leopold-Franzens-Universitaet Innsbruck Institut fuer Kommunikation im Berufsleben und Psychotherapie
Schoepfstrasse 3 6020 Innsbruck, Austria
“Cooperation is important for us, but sometimes it is like playing with fire.” A qualitative analysis of the
Title: development and management of social capital in technology orientated start-ups within an academic business
In recent years academic business incubators have been increasingly recognized as important for transferring
academic expertise to private business. In this paper we focus on the process of social capital development and
the practices entrepreneurs use to manage their social capital. Firstly, we explore the dynamic development of
social capital of start-ups and the areas of tension the entrepreneurs are facing. Furthermore, we analyse the
role an academic business incubator plays in this development. Using a qualitative social network approach we
present results from case studies of an Austrian academic business incubator and its technology oriented
start-ups. In general, results offer support for the existing view that start-ups possessing a rich network regarding
social capital have been far more successful in view of number of employees and sales. We argue that in the
start-up process social capital cannot be seen as static. Social capital changes in different stages of the start-up,
depending on the resources currently needed. The success of the entrepreneur depends on their ability to
identify the changing business needs and to quickly adjust their social networks to the resources required. Due to
the persistence forces of social capital, entrepreneurs are facing problems on the relational and cognitive
dimensions, such as the conflict between academia and business, and cooperation and competition. Moreover,
results suggest that a principal role of the academic business incubator is in providing credibility and reputation
for entrepreneurs – especially young ones. Finally we discuss these results with regard to the management of
social capital for entrepreneurs and academic business incubators.
Session: Social Capital
Keywords: social capital , qualitative network , entrepreneurship , business networks , inter-organizational
Accepted: Yes

35 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Johannes M. Bauer
Quello Center for Telecommunications Management and Law 409 Communication Arts and Sciences Michigan
Authors: State University East Lansing, Michigan 48824, USA
Volker Schneider
University of constance, Politics and management, Universitaet Konstanz, Box D 81, Konstanz,
Baden-Württemberg 78457, Germany
The Complexity of Associational Systems: Networks of Cooperation and Competition among Business
Associations in the U.S. Information and Communication Technology sector.
In political science, industrial economics and political sociology interest groups and associations are often seen
as pure lobbying mechanisms or pressure groups that illegitimately inject private interest into processes of public
policy making. Such perspectives often suggest that societies would be better off without private interest
interference. However, on the other hand, there are theories like “neo-corporatism” or “private interest
governance” which point to the important contributions of interest associations in the governance of modern
societies. Such theories derive the capacity for public governance functions from certain structural characteristics
of associational interest representation in policy processes (e.g. hierarchical set-up, monopolisation of interest

Both perspectives are overly simplistic: Whereas the first is based on functional simplicity, underestimating the
multi-functionality of associations, the second is based on structural simplicity, exaggerating simple orders and
uni-dimensional structures (e.g. juxtaposing cooperation or competition). Compared to the pluralist thesis and the
corporatist antithesis, this paper portrays a new synthesis based on complexity theory. In this perspective
associations are involved in lobbying and in self-regulation. They interact in partially competitive but also
Abstract: cooperative relationships.

The paper applies this perspective to national trade associations in the information and communication
technology (ICT) sector of the United States. Analysis is based on data of a focal set (N=19) of business
associations which have been identified by an expert panel based on influence reputation. The paper describes
and analyses how these organizations perceive challenges of competition, convergence, and globalization (three
important developments affecting the ICT industries), and how they cope with environmental pressures. A major
focus is how the various trade associations compete and cooperate with each other and how they exchange
information and other resources. Through network analysis of data on contact relations, information exchange,
interest similarity, overlapping membership, and domain competition we are able to identify collective actors and
key associations playing an important role in the horizontal coordination and regulation of this heterogeneous
sector. In a further step the paper also analyses influence and lobbying strategies with regard to specific resource
allocation profiles and the targeting of major policy makers at the domestic level of the USA as well as
international organizations such as WTO, WPO, EU and OECD playing an important role in the international
regulation and standardization of this industry.
Session: Policy Networks and Governance
associations , collective action , economic networks , interest representation , self-regulation , coordination ,
governance , policy network
Accepted: Yes

36 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Hélène Bayard Çan

Çukurova University, Department of French Language Teaching, Adana, Turkey Paris X Nanterre University,
Department of Ethnology, Paris, France
Title: A Neighbor Network in Urban Turkey
Neighbor relationships occupy a fundamental place in Turkish daily sociability and represent both strong and
weak ties. However, they have been little studied and have never been approached by the way of social network
This paper studies the structural characteristics of a urban neighbor network in South Turkey (Adana) by
exploring the connections inhabitants develop within neighbor community. I try to analyze in which way the
studied neighbor network is related to gender but also age, occupation and presence of relatives within the
neighbor network. My research is based on ethnographic field observations as well as on 40 interviews
conducted with inhabitants of 4 apartment buildings which permitted to know who neighbored with whom and
how. The results of this qualitative data led to a 102 nodes valued network analyzed with Ucinet and Netdraw.
The results of the research show that family ties play an important role to constitute the neighbor network and
enlarge it. Furthermore, housewives entertain the most neighbor relationships, while men or working women tend
to have more superficial neighbor relationships and with a minor number of neighbors.
Session: Neighborhood Communities
Keywords: neighborhood community , turkey , social network , ucinet , qualitative research
Accepted: Yes

37 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Mariano Belinky
Authors: Dept. Llenguatges i Sistemes Informàtics Universitat Politècnica de Catalunya Barcelona, Spain
Bruce Kogut
Professor of Strategy Insead Business School Fontainebleau, France
Title: Comparing boards and ownership affiliation networks over time and across countries
Board of directors have received considerable attention in the social networks literature. Following recent work on
board and ownership affiliation networks, they have become widely used as test cases for several social
networks features. Since these networks present different sizes and degree distributions across countries and
temporal cross sections, we face the issue of how to analyze and compare them. One of the features featured in
these comparisons is the networks' degree distribution, which tends to be arbitrary. However, recent approaches
in the literature use classic distributions to generate base cases from which to compare across networks.
Moreover, the common approach is to concentrate on the one-mode projection instead of the full bipartite
In the present work we summarize different methodologies used in the literature to compare board networks over
time and across geographical boundaries. We take five different countries, estimate their network statistics, and
compare their boards of directors
interlocking. This analysis provides a comparative analysis of the "small-worldness" of these networks. We also
look at their homophily, also known as assortativity, and community structure to understand better the underlying
rules that might explain the differences among these comparisons. Moreover, we look at a second group of
networks, based on ownership ties, and analyze their particular features as well as using them to supplement the
boards analysis.
Session: Inter-Organizational Networks
Keywords: affiliation networks , interlocking directors , standardization , network structure , network comparison
Accepted: Yes

38 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Elisa Bellotti
Authors: elisa.bellotti@unicatt.it
Università Cattolica del Sacro Cuore, Sociologia, via brunelleschi 4, Milano, Mi 20146, Italy
Title: Friendship Networks of Single Youths
The paper presents the results of research on friendship networks of single people. The research has been
undertaken using both social network analysis tools (egonetworks) and qualitative interviews, on a
non-representative sample of 23 heterosexual singles of Milan, Italy, aged between 25 and 35, half male and half
female. Single youths has been portrayed as a generation who decentres sexual and romantic relationships and
Abstract: refuses to commit: this paper demonstrates the need for new analytical concepts in order to describe the
contemporary transformations of intimacy. Friendship is a source of emotional, social, material and economic
support, that sometimes can offer a valid alternative to the traditional couple relationship. Four main structures of
friendship networks have been found: the small clique, the group, the core-periphery structure, and the
contextualised components.
Session: Qualitative Network Studies - II
Keywords: friendship networks , qualitative network
Accepted: Yes

39 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Corinne Bendersky
Authors: University of California, Los Angeles, Management, Los Angeles, CA, USA
Neha Shah
UCLA Anderson, Los Angeles, CA, USA
Title: Mobility and stability in status: Implications for individual outcomes
Status-- social evaluations associated with influence and resource acquisition across individuals-- has drawn
significant attention in the organizational literature. The great majority of scholars who study status consider
implications of a stable hierarchy. Much of this research suggests those who hold high levels of status in
networks achieve better workplace outcomes, such as higher performance evaluations, than those with low
levels of status. Yet an individual’s level of status may fluctuate over time, and researchers rarely consider
outcomes associated with changes in status. Thus, this paper examines outcomes associated with status
mobility, gaining and losing status, compared to status stability, maintaining high or low status, over time. In a
Abstract: longitudinal study of 108 MBA students who were organized in study groups for a 10-week class quarter, we
consider how individual performance and satisfaction vary across individuals who gain, lose, and maintain
different levels of status and influence, measured as changes in indegree centrality. We also control for individua
traits (i.e. age, race, gender, work experience) that have been shown to affect status perceptions. Additionally,
we consider implications associated with ego’s agreement or disagreement with alters’ perception of changing
status and influence. Preliminary results suggest status and influence mobility and stability affect performance
and satisfaction quite differently. Final results may include other dependent variables, such as group
performance and individual perceptions of group potency and cooperation.
Session: Network Dynamics
Keywords: social influence , longitudinal
Accepted: Yes

40 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Laura Bernardi
Max Planck Institute for Demographic Research Independent Research Group Culture of Reproduction Rostock,
Authors: Germany
Clementine Rossier
Institut National des Etudes Démographiques, Paris, France
Title: Entry into parenthood, social influence and social cohesion
The paper presents a mixed-method research design to study the role of social interaction with peers and kin in
the transition to parenthood in Germany and France. We investigate in which way the different configuration of
young couples’ social networks affect fertility intentions and behavior. To detangle theses complex processes, we
collected semi-structured interviews with a purposeful quota sample of men and women of parity 0 and 1. The
originality of our data consists in the use of an adapted version of the hierarchical mapping procedure employed
successfully in social psychology as an open stimulus to assess what kind of relationships are relevant for fertility
decision-making. Respondents are invited to define their current social networks by completing a network chart
and a grid and commenting on their relationships with each network member mentioned in the chart. The network
chart represents a mixed method data collection tool which is conceived to gain information to be analyzed
qualitatively and quantitatively. First, we produce a rich description of the process and the mechanisms of social
influence by a applying sistematic content analysis to the data produced by the chart and by the biographic
narrative questions. Second, we use the information of the chart in combination with that of a more classic
network grid and questionnaire, to asses whether social influence mechanisms vary in differently configured
Session: Mixed Methods: Combining Qualitative and Quantitative Approaches
Keywords: social networks , mixed methods , fertility , social influence , social learning , family
Accepted: Yes

41 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Stefan Bernhard
Doctoral Program "Markets and Social Systems in Europe", University of Bamberg Otto-Friedrich Universität,
96047 Bamberg, Germany
Title: Semantic Lobbying in European Policy Fields – Networking in Social Inclusion Policy
Network analysis is surprisingly absent in scientific literature on „new modes of governance“ (Heritiér 2002) in
Europe. This contribution therefore aims at showing the theoretical and empirical potential of a network
perspective on European governance with regard to the so-called Open Method of Coordination in the area of
social inclusion policies. In the absence of “sanctioning power” (Scharpf 2000) civil society actors are expected to
contribute substantially to the coordination of national policies in these processes: they are supposed to control
Member States with respect to their policy aims and they are in the position to contribute expertise. It is argued
that civil society actors systematically use their strong position in new modes of governance by knotting a tight
network among one another and towards European Institutions. Against this background they pursue different
strategies of “semantic lobbying”, i.e. trying to influence the way policies are framed at the European level.
The argument has three parts: (1) Bourdieu’s theory of social field constitutes the starting point of analysis. In a
first step the theoretical potential of relating his understanding of capital, field and hegemonic struggles to
network analysis is developed along their respective understandings of “structures” and “capital”. (2) A second
step outlines the Open Method of Coordination as formally institutionalised structure of the European field of
social inclusion policy. The passage illustrates the expectations towards civil actors and highlights the networking
opportunities within the process. (3) Finally several strategies of semantic lobbying at EU level are identified
ideal-typically. Depending on their possession of different types of capital, i.e. cultural, economic, social and
political capital (Münch 2001), civil society actors follow different ways of achieving symbolic capital, i.e. the
capacity to define policy framing (Bourdieu 1983).
The investigation implements a stricktly “relational approach” (Emirbayer 1997) using a qualitative network
analysis of network maps polled in 20 expert interviews (Hollstein 2006; Kahn and Antonucci 1980).
Session: Civil Society Networks
Keywords: civil society , eu policy making , knowledge networks , policy network , social inclusion , qualitative research
Accepted: Yes

42 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Kathryn Berzins
Authors: K.M.Berzins@clinmed.gla.ac.uk
University of Glasgow, Public Health and Health Policy, 1 Lilybank Gardens, Glasgow, Scotland G12 8RZ, UK
Title: Social networks of people using mental health services compared to the general population
People using mental health services can experience exclusion across the whole of social and economic life,
figures from England and Wales in 2000 showed that people with a psychotic illness are three times more likely
to be separated or divorced than those without. They are also twice as likely to live alone (ONS, 2000).
The aim of this research was to explore this exclusion by focusing on the composition of the social networks and
perceived levels of social support reported by people using mental health services, in comparison with a general
population sample.
Interviews were carried out across Scotland with 200 people using mental health services and 1015 members of
Abstract: the general population.
The analysis compared the two groups with regard to composition of social networks and levels of social support.
It showed differences in perceived levels of social support and differences in both network size and composition.
It further compares the roles played by mental health professionals for people using mental health services that
are fulfilled by family and friends in the general population.

Office for National Statistics (2000) Survey of Psychiatric Morbidity Among Adults Living in Private Households.
London, Office for National Statistics.
Session: Social Support
Keywords: family , health care providers , social inclusion , social support , personal networks , friendship networks
Accepted: Yes

43 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Paula Bialska
Authors: Department of Sociology, University of Warsaw, Warsaw, Poland
Dominik Batorski
Department of Sociology, University of Warsaw, Warsaw, Poland
Title: Trust Networks: analyzing the structure and function of trust
The roots of trust and what enables trust to take place between individuals is generally difficult to calculate and
measure. Within this study, we make use of a specific online social hospitality network known as
Couchsurfing.com, which, because of it’s character, enables us to measure the emergence and function of trust
in a more explicit way. Essentially, trust is a necessary factor within such a network – where 156,000 users
create on-line profiles and ties with other users in order to travel to foreign places while residing in the private
home or “couch” of other members of the network. The network is based on a unique friend-link system which
shows the degree of trust between diads explicitly. By studying a network where trust plays such a crucial role,
we can more accurately identify the way trust emerges and functions between individuals.
This analysis is based on anonymized whole network data provided by the administrators of Couchsurfing.com.
Through this study, we distinguish if factors such as homophily, the duration of a relationship, and the number of
common friends create a higher level of trust. We also focus on the difference of trust levels between those
contacts created online versus those created offline, and if an individual’s frequency of outgoing high trust links
makes him/her more trustworthy in the eyes of others. Additionally, we compare which of the previous factors are
more important in the trust exchange.
Session: Networks and Trust
Keywords: trust , trust in networks , friendship networks , exchange networks , homophily , social relationship
Accepted: Yes

44 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Yanjie Bian
University of Minnesota, Dept of Sociology, Dept of Sociology, Univ of Minnesota, 267 19th Ave South,
Authors: Minneapolis, MN 55455, USA
Xianbi Huang
The University of Queensland, UQ Social Research Centre, Level 4, GPNorth 3 (Bldg. 39A), Brisbane,
Queensland 4072, Australia
Title: Favor, Matching, and Structural Holes: Network Effects on Wage Income in China
How do social networks matter for labor market outcomes? Recent critiques identify design deficiencies of
previous empirical studies that used probably misguided measures of tie strength and contact characteristics as
predictors of the network effect on wage income. Building on network theories and the research of transitional
China, in this paper we propose three causal mechanisms--favor, matching, and structural holes--by which
network contacts transmit information and influence that in turn result in higher wage income of Chinese workers
who used contacts to find jobs than those who did not. In a 1999 random sample of the general population of five
Chinese cities (N=4350), 59% of the respondents found their current or last jobs through social contacts who
Abstract: provided job information (32%) or more concrete favors (38%), or both (21%). The concrete favors include
delivering applications, face-to-face recommendations, setting up informal interviews, and the like, and they have
an immediate impact on a higher initial salary for favor-receiving workers. In addition, these favors and the
information transmitted both have a long-term, positive impact on wage income by ways of assigning the workers
into jobs that better match worker qualifications to the job requirements of skill training and work experience, and
into positions that give the occupants a greater diversity of organizational and market connections to earnings
opportunities. The significance of these network effects is discussed within the institutional context of China and
Session: Social Capital
Keywords: social capital , informal networks , economic networks
Accepted: Yes

45 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Claire Bidart
Authors: LEST-CNRS, Sociology, 35 Avenue Jules Ferry, Aix en Provence 13626, France
Johanne Charbonneau
INRS-Urbanisation, culture et société, 385, rue Sherbrooke est, Montréal, H2X 1E3, Québec, Canada
Title: The contextual name generator: a good tool for the study of sociability and socialization
The debate upon the relative validity, power, limits and relevance of the different name generators comes along
the development of the social network studies. The core questions are: what do they respectively refer to? What
are they supposed to construct, for what research question? Some procedures tend to choose a precise target
with a unique name generator that may synthesize a crucial point. Others prefer to use series of different name
generators, in order to gather names referred to diverse spheres of social life. In this case the various name
generators are often built with heterogeneous logics, and often remain uncompatible.
Is it possible to standardize a procedure enough to overcome these limits and keep the comparisons possible?
We discuss here some specificities and advantages of a new kind of integrated name generator, the “contextual”
name generator, which was developed in a longitudinal qualitative panel study that started in France in 1995 and
was also conducted in 2005 in three different projects in Quebec. This tool is not the juxtaposition of independent
name generators, as we are used to; it combines their respective advantages in a real integrated and systematic
procedure and allows going through a wide range of areas, scales, social conditions, qualities of ties, etc. This
name generator gives access to a great diversity of information that allows to combine sociability and
socialization questions. It thus seems to be a relevant tool, especially for sociologists.
To know more about these surveys:
http://www.lest.cnrs.fr/article.php3?id_article=375 (french) http://www.lest.cnrs.fr/article.php3?id_article=376
Session: Qualitative Network Research: Data Collection
Keywords: data collection , information diversity , methodology , personal networks , qualitative research , social contexts
Accepted: Yes

46 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Walter Bien
Authors: German Youth Institute, Nockerstrasse 2, Munich D-81541, Germany
Jan Marbach
German Youth Institute, Nockerstrasse 2, Munich D-81541, Germany
Family Networks over Time: Dynamic Interrelations between Family Development, Structural Composition and
Social Capital Output
The temporal dynamic of an ego-centered network, as measured at distinct points in time, can be depicted using
set-theoretical tools, like intersection, complement, union etc. This operation is premised on identifying members
of temporal separated networks of a constant Ego as identical or different. When applied to a large body of panel
data the procedure can turn out rather cumbersome, but it seems worthwhile to accomplish. There are two major
sets of questions possible to be answered using information about the stable core and varying subsets of a
personal network over time. First: Considering the network a dependent variable one may ask how and to what
extent respondents´ bygone life trajectory is responsible for the change observed. Is, e.g., the network´s
composition of stable and shifting building blocks related to changes of a respondent´s living arrangement or
concurrent influences as gender, age, education, moves of parity, occupation, residence etc.? A second set of
questions may be even more interesting: Employing network change now as an independent variable one may
ask how and to what extent the structural change of an ego-centered network is effecting its quality as Ego´s
resource, i.e. Ego´s social capital. For example, does a large stable network core strengthen "strong ties" in the
sense James S. Coleman delineated the social capital output of a personal network? On the other hand, do
"weak ties", as Mark Granovetter defined them, benefit from a network mainly consisting of alternating members
over time? Is an association, if any, mediated by attributes of Ego e.g. state of his or her family development? Or
properties of the network members e.g. who establish the core of the network , or who show up in its volatile
margins respectively?
In our presentation we will pick up some of these questions starting with the first block but focussing on the
second one, the relationship between structural change of personal networks and their social capital output. Our
data base is a panel ingredient of the German Family Survey. The panel study covers a 6 year interval starting
with a baseline survey in 1988 that captured more than 10,000 interviews of (then Western) Germans aged 18 to
55. A second surge of interviews was finished in 1994. Findings rest on 5000 panel interviews in the area of
former West Germany. Ego-centered network data were gathered by means of 12 name generators and 7 name
Session: Social Capital
Keywords: dynamics on networks , family , egocentric networks
Accepted: Yes

47 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Stephen Bird
Authors: sbird@bu.edu
Boston University, Political Science, 33 Parkman Street, Dorchester, MA 02122, US
Title: Collaborative Social Networks: Collaboration, not Information Diffusion, in the Classroom
This paper investigates the relationship between social networks and grade performance for college students.
Quasi-experimental results in a large, first-year introductory course (n = 294 students) demonstrate a correlation
between collaborative academic social networks and grade performance. They further imply that weak ties have
substantive importance even in collaborative, interactive learning processes. While it is to be expected that socia
networks would have a positive net effect on grades, the impact of social networks on grades is substantial.
Second, this analysis illustrates the importance of collaboration (as opposed to information diffusion) on the
learning process. Third, while weak ties and structural holes theory have strong explanatory power for social
networks, collaborative interactions could be more strongly reinforced in situations where strength or intensity of
social interaction are higher but this is not the case. Instead, weak tie and structural hole theory continue to
demonstrate strong effects even when the interactions are collaborative rather diffusion oriented. Lastly, the
benefits of such networks are limited beyond certain levels of collaboration. The results bolster much of the
theoretical implications of the Communities of Practice and Social Learning literature.
Session: Knowledge Networks
centrality , collaboration , diffusion , experiments , individual performance , knowledge transfer , innovation ,
peer education
Accepted: Yes

48 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Martin Bliemel
Authors: mbliemel@sfu.ca
Simon Fraser University, Busines Admin, 515 West Hastings St., Vancouver, BC V6B 5K3, Canada
Title: Relative Relational Embeddedness: Towards a Normal Integration of Resources and Relations
This paper develops a framework for analyzing an organization’s capacity to apply its competitive resources in a
network context. The framework is based on an integration of the internally oriented resource based view (RBV)
and the externally oriented social network (SN) perspective. The integration of these perspectives addresses the
paradox that while internal and external resources can be combined to complement each other to creave value to
those involved in the relationship, firms are limited in their ability to develop any combination of internal and
Abstract: external resources. The framework is based on considering relational embeddedness relative to the firm’s
available resources, coined relative relational embeddedness (RRE) here. Examples are given for four scenarios
in which firms balance two dimensions: firstly balancing internal and external resource development, and
secondly balancing strong and weak ties in the external network. In lieu of empirical data, an extension of a
related operationalization of structural embeddedness is provided with which scholars may put the framework to
the test.
Session: Innovation
Keywords: embeddedness , inter-organizational networks , business networks , egocentric networks , entrepreneurship
Accepted: Yes

49 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Annette Bobrik
Authors: Technical University Berlin, SYSEDV, Franklinstrasse 28/29, Berlin, Berlin 10587, Germany
Matthias Trier
Technical University Berlin, SYSEDV, Franklinstrasse 28/29, Berlin, Berlin 10587, Germany
Title: Visually tracing keyword diffusion in social networks
The dissemination of topics in social networks is of interest in the domains of viral marketing and topic trend
detection. In our contribution, we present an method which is combining algorithms from social network analysis,
dynamic animation of network evolvement, and longitudinal text mining to allow for a visual observation of how
(collections of) keywords, i.e. product names or issues, emerge and spread in virtual communication networks.
Using datasets of Enron and online discussion boards, we demonstrate how this approach can support the visual
observation of a collection of search terms in an evolving communication network. It allows insights into the origin
of a topic, i.e. whether it has first been adopted in a single locus or at multiple places. Further, it can be traced,
how it was disseminated through the actor’s network and where groups found each other over time. Although
such longitudinal keyword analysis in networks has limited potential to explain complex issues, we argue that
together with measuring actor activity and the number of keyword occurrences, this can help to generate quick
insights into topic trends and lifecycles in communication networks and thus is able to augment social network
analysis with a content oriented perspective.
Session: Communication Networks
Keywords: content analysis , visualization , network dynamics , network evolution , changes in networks
Accepted: Yes

50 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Örjan Bodin
Authors: Stockholm University Dept of Systems Ecology Stockholm, Sweden
Beatrice Crona
Stockholm University Dept of Systems Ecology Stockholm, Sweden
Management of natural resources at the community level – exploring the role of social capital and leadership in a
rural fishing community
Social capital and leadership are important factors affecting natural resource management. This study explores
aspects of social capital and leadership in a rural fishing community to seek explanations for why collective action
for resource regulation has not occurred despite declining fisheries.

Results show relatively high levels of social capital as quantified through social network analysis, but low
willingness to report rule-breaking. Furthermore, links to external agencies have previously been shown as
crucial for leaders to enable communities to make use of their social capital, and key individuals identified here
were shown to possessed links to several external agencies but only few links to financial institutions and
markets beyond the trade of fishing-gear. These findings, may, individually or in combination, explain lack of
common initiatives in the village to deal with the overexploitation of fisheries. However, additional hypotheses are
also put forth. Homogeniety among key persons may have lead to the low levels of problem internalization and
recognition of changing ecological conditions. Secondly, structural characteristics of the social network among
key individuals reveal one person to occupy a very central position. The ability of the community to collectively
organize is thus, for good or bad, largely dependent on one very central actor.
Session: Social Capital
social capital , access to knowledge , centrality , informal networks , network structure , natural resource
Accepted: Yes

51 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Michal Bojanowski
Authors: m.j.bojanowski@fss.uu.nl
Utrecht University, ICS/Sociology, Heidelberglaan 2 (vanUnnikgebouw), Utrecht 3584 CS, Netherlands
Title: Co-evolution of networks and behavior under heterogeneity
Most of the existing theoretical contributions to understand mechanisms of co-evolution of networks and
individual behavior assume that actors are homogeneous. The consequences of relaxing this assumption are not
yet fully understood. Under which conditions will the differences between actors result in higher segregation
levels than in the homogeneous case? In this paper we study the interrelated dynamics of social networks and
behavior when actors' interests differ. As a framework for analysis we propose a baseline model in which actors
simultaneously choose their behavior and manage their personal relations with others. The population of actors is
composed of two types and interactions are modeled with asymmetric two-person games. The heterogeneity is
represented by three elements:
(1) the degree to which actors' interests behavioral options differ,
(2) the severance of "miscoordinating"
(3) complementary or substitutable character of relations with actors of the other type.
To address the posed problems and evaluate the role of the three above mentioned components we employ
analytical and computer simulation methods. First, we look at what kind of stable network architectures emerge if
actors actively try to improve their position by making behavioral and relational choices. Second, we investigate
the link between the level of divergence in the population in terms of types and behavior, and the level of network
segregation in the emerging networks.
Session: Mathematical Models
Keywords: coevolution , network dynamics , conflict , simulation study
Accepted: Yes

52 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Phillip Bonacich
Authors: bonacich@soc.ucla.edu
University of California, Sociology, Department of Sociology, Los Angeles, California 90095, USA
Title: A New Measure of Centrality
I propose a new family of measures of centrality that is based on network dynamics and that permits varying the
“depth” into the network to which it is responsive. The effects of patterns of influence or communication on
individuals could be expressed as a set of differential equations involving individual characteristics (status,
information) x, change in these characteristics x’, the adjacency matrix A, and time, t.

x’(t) = Ax(t).
The solution to this set of equations is:

x(t) = exp(At)x(0)

The matrix exp(At) shows the time-dependent influence of individuals, where they first influence those near
themselves, and then, as time increases, those more and more remote in the network.
Session: Formal Methods and Statistical Models
Keywords: social network , centrality
Accepted: Yes

53 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Matthew Bond
Authors: m.bond@kent.ac.uk
University of Kent SSPSSR Cornwallis Building Canterbury Kent CT2 7NF
Title: Social networks and corporate political action: the case of Business for Sterling
In this paper I study the socio-economic correlates of the decision of British businessmen to join Business for
Sterling. I study the 3992 directors of the 500 largest corporations in Britain in 97-8. I find that social networks
Abstract: created through clubs and interlocking directorates are associated with the decision to join along with being
extremely wealthy. This paper shows how social network measures can be used to operationalise many concepts
in elite research.
Session: Business Networks
Keywords: interlocking directors , collective action , cultural communities
Accepted: Yes

54 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Johannes Boshuizen
Authors: University of Twente, School of Management and Governance, Enschede, The Netherlands
Anne Van der Veen
University of Twente, School of Management and Governance, Enschede, The Netherlands
Title: Interfirm networks and the success of regional clusters
Regional clusters have attracted increasing interest from researchers, governments and consultants during the
last years. One reason for the popularity of the cluster concept is the success of famous regions such as Silicon
Valley. The central idea is that interactions among firms in a cluster generate crucial benefits for firms and are
necessary for the success of a cluster. Governments all over the world try to stimulate the development of their
own clusters, for example by supporting cooperation between firms. Given these efforts, it is surprising that the
amount of empirical research on interactions within clusters is limited.

According to recent theories, interactions within a cluster give firms several local advantages. Firms can rely on a
diverse local knowledge pool and can exchange information of all kind. Informal relations are seen as important
means to establish these exchanges. Easy face-to-face contact and trust keep transaction costs low. We argue
that social network methods and theory provide a promising approach to research various types relations within a

The objective of our research is to map the network among firms using multiple sources of data. This paper
draws on data on membership of business associations. We present preliminary results of the analysis of the
affiliation network among high tech firms in the Dutch region of Twente.
Session: Inter-Organizational Networks
Keywords: affiliation networks , inter-organizational , regional clusters , innovation , embeddedness
Accepted: Yes

55 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Moses Boudourides
University of Patras, Mathematics, 265 00 Rio-Patras, use-->Meilihou 58, 264 42 Patras, Greece, Patras,
Peloponnese-Achaia 264 42, Greece
Title: The Relational Ontology of Social Network Theories
An always important philosophical investigation concerns the question “What are Beings?” that Heidegger
regarded as the refrain of Western metaphysics. Social network theories favor a relational approach in answering
this question. In the social network context, the traditional philosophical ontology of individuality (Kant, Nietzsche,
Pierce etc.) shifts towards an ontology of relational processes that operates as a duality between the micro and
the macro-level, i.e., towards something which amounts to a structural relational ontology of individuation.
However, there are more philosophical answers to the question o individuation. Our purpose is to examine here
the work of two contemporary French philosophers and to discuss how such philosophical ontological
perspectives could possibly extend the scope of social network theories. The first one is Gilbert Simondon, who
Abstract: had advanced a project to know the individual through processes of individuation rather than the other way
around. His theory of individuation, starting from a ‘pre-individual’ stage and passing through ‘transduction’ in a
‘meta-stable’ environment, was an import influence to the thought of Gilles Deleuze. The second philosopher we
are discussing here is Alain Badiou, who has tried to reconcile a notion of the subject within a relational ontology
throughout a number of propositions in one of his major works entitled ‘Being and Event.’ Badiou’s philosophy is
based on the axioms of set theory that he essentially identifies with the language of all ontological investigations.
Furthermore, in his recently published second volume of ‘Being and Event’ (Livre IV des Logiques des mondes),
Alain Badiou explicitly discusses relationships, as a meaningful way to frame an ontological discussion around
the question of the existence of structural patterns.
Session: Ontology and Philosophy of Networks
Keywords: philosophy , ontology , structural approach , relational approach
Accepted: Yes

56 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Mohamed Boukhris
University of Lugano faculty of Economics Switzerland mohamed.boukhris@lu.unisi.ch
Authors: Alessandro Lomi
University of lugano faculty of Economics Switzerland lomia@lu.unisi.ch
lei Wang
University of Lugano faculty of Economics Switzerland lei.wang@lu.unisi.ch
Network structures, behavior and performance: preliminary results from a longitudinal study of mergers and
acquisitions in the international electricity industry
As deregulation of key industries continues to gain momentum around the world, the electricity sector is rapidly
changing from a series of local government controlled monopolies to a competitive global industry through waves
of mergers and acquisitions. This large-scale economic change is brought about mainly by the erosion of
boundaries around established corporate actors, conventionally defined industries and – in the longer term –

In this paper, we present preliminary results of a longitudinal study of the network dynamics of mergers and
acquisitions in the international electricity industry during the period 1994-2003. The nodes in our networks are
Abstract: represented by individual countries and the edges by foreign investment flows among countries induced by
acquisition activities among companies. We are interested in understanding the local relational building blocks of
global industry structure and how network-based processes at the same time affect and are affected by relevant
variables such as price and production capacity. We explore these issues by specifying actor-oriented models
developed for the longitudinal statistical analysis of social networks.

We focus on mergers and acquisitions in the international electricity industry because this industry provides an
almost ideal illustration of how dyadic network ties induce global network structures that enable and constrain the
behavior of business firms, industries and national economies across multiple levels of analysis
Session: N/A
Keywords: network dynamics , coevolution , merger and acquisitions
Accepted: Yes

57 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jeffrey Broadbent
Authors: broad001@umn.edu
University of Minnesota, Department of Sociology, 909 SSB, 267 19th Ave. South, Minneapolis, MN 55455, USA
Building on a Rock: Modifying the Labor Policy Network Survey Instrument to Address Global Climate Change
Politics, Cognitive Fields and Institutional Effects
The Compon (Comparing Climate Change Policy Networks) project builds on the solid design of the survey
instrument used for the Comparing Policy Networks (CPN) study of the labor politics domain in the US, Germany
and Japan (Knoke, Pappi, Broadbent and Tsujinaka 1996). However, the Compon project is studying a different
domain, climate change, in an era of increasing globalization, with a focus on the impacts of scientific knowledge
from the IPCC (and other sources) on national policy counter-measures. This requires an expanded set of
theoretical frames, including different relational networks and new ways of capturing the flow and acceptance of
Abstract: more narrowly specified types of beliefs about climate change. In addition, the range of cases will expand from
the leading industrial societies to a wider array of societal types, including China, Taiwan, South Korea, Japan,
Canada, the US, Brazil, Peru, England, the Netherlands, Germany, Sweden, Austria, Greece and Russia, and
possibly other cases as well. These expanded foci require an enhancement of the CPN design to operationalize
these new perspectives. The paper will examine the new requirements and make suggestions as to what specific
new questions and other design features are needed in the common survey instrument to maximize the collection
of usable data on these new questions and issue.
Session: Policy Networks on Climate Change
policy network , political networks , environmental networks , climate change policy , environmental
performance , inter-organizational networks , international networks , kyoto protocol
Accepted: Yes

58 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Nathaniel Bulkley
Authors: nbulkley@uiuc.edu
University of Illinois - Urbana Champaign, NCSA, 20 Parsley, Savoy, IL 61874, USA
Title: Using Electronic Trace Data in Social Network Research: Issues and Opportunities
Social network studies increasingly incorporate trace data from media such as email, instant messaging, cell
phones and Web sites. Direct measurement offers opportunities to study social networks on a larger scale, at a
finer level of granularity, and in ways that address concerns over informant inaccuracy in surveys. However, after
obtaining electronic data, researchers face the challenge of interpreting traces subjects left for reasons that are
often unclear under conditions that are at best partially understood.

In this presentation, we draw principles for analyzing electronic trace data from general theories of measurement
as they have been applied to non reactive measures. We illustrate the application to social network analysis with
examples from our ongoing research, which relates measures of email activity to individual performance in the
context of executive recruiting.
Topics include: conceptualizing, defining and validating measures; constructing analyses to differentiate between
correlates of direct measures; and the importance of planning secondary data collection to facilitate subsequent
testing of plausible competing hypotheses. We will also discuss implications of direct measurement for
interpretation and future development of existing social network theory.

Electronic data sources and computational techniques for analysis represent new instruments for conducting
social network analysis. Scientific breakthroughs often involve new instruments, methodologies for making sense
of observations enabled by these instruments and the creative refinement of theory to incorporate new results.
Adopting this historical perspective, we see general theories of social science measurement offering a useful
vantage point for pursing research that capitalizes on capabilities of these new instruments.
Session: Network Evaluation
methodology , e-mail database , method triangulation , nonreactive measures , performance measurement ,
inductive theory-building
Accepted: Yes

59 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Marlene Burkhardt
Authors: burkhardt@juniata.edu
Juniata College, IT & CS, 1700 Moore Street, Huntingdon, PA 16652, USA
Title: Longitudinal analysis of technology social networks and cell phone attitudes survey data
Longitudinal research was conducted to assess the relationship between cell phone attitudes and technology
social networks. A cell phone attitudes survey was administered to sports teams within a small liberal arts college
and network data was collected on cell phone directory, text messaging, and instant messaging partners within
the sports team. The data was collected at two points in time, approximately one year apart. As hypothesized,
findings indicate that the cell phone directory network and text messaging network are positively related to
similarity in beliefs that cell phones are helpful in daily life, easy to use, increase productivity, and are worth the
money. Several findings were observed to be significant at both points in time. The Time 1 cell phone network
Abstract: was significantly related to the Time 2 similarity in productivity attitudes lending some support for the hypothesis
that our beliefs regarding productivity are influenced by those who we list in our cell directories. The Time 1
instant messaging network was related to the Time 2 ease of use and helpful similarity directories. This implies
that we are influenced by those with whom we instant message. Cross-lagged correlational analyses indicate that
the Time 2 cell phone directory network data (in comparison to Time 1 data) is a better predictor of Time 1 helpfu
and worth it matrices, building some support for the belief that we talk to others who are similar to ourselves.
Supplementary analyses found Facebook network data to be a very poor predictor of social influence. Research
findings have relevant implications for marketing opportunities.
Session: Innovation
Keywords: social influence , social network
Accepted: Yes

60 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Vincent Buskens
Authors: Utrecht University, Sociology, Heidelberglaan 2, Utrecht 3584 CS, The Netherlands
Rense Corten
Utrecht University, Sociology, Heidelberglaan 2, Utrecht 3584 CS, The Netherlands
Title: Co-evolution of conventions and Networks: An Experimental Study
This paper presents an experimental study on the emergence of conventions in dynamic
networks. Conventions are modeled as coordination games in which actors can choose both
their behavior and their interaction partners. First, we address the question how the
extent to which a single convention (rather than several) can be reached depends on the
initial structure of the network. Secondly, we study how efficiency of resulting
conventions (the extent to which Pareto-optimal conventions are chosen) depend on this
initial structure. Third, we study the effects of the availability of information on
behavior of others on the emergence of conventions. Predictions come from a
game-theoretic model in which actors play coordination games on a dynamic network. In
order to derive accurate predictions for the experiment, we simulate our experimental
conditions in a computer simulation analysis of this model. The simulation results
predict that 1) the higher the density of the initial network, the more likely it is that
the network will converge to one convention, 2) the higher the density of the initial
network, the more the efficiency in the initial network will determine the efficiency of
stable states and 3), results suggest, somewhat counterintuitively, that less information
leads to a higher likelihood that a single convention will be reached. These hypotheses
are tested in controlled, computer-guided experiments in which groups of eight actors play
coordination games while also choosing their interaction partners. We analyze both
macro-level outcomes and individual behavior. The results provide mixed support for the
Session: Network Dynamics
actor-driven modeling , game theory , homophily , coevolution , coordination , network dynamics ,
experiments , simulation
Accepted: Yes

61 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Carter Butts
University of California, Irvine, Department of Sociology, Institute for Mathematical Behavioral Sciences, Irvine,
CA 92697, USA
Title: Network Inference from Unstructured Sources
Prior work by Romney, Weller, and Batchelder (1986), Batchelder and Romney (1988), and Butts (2003) has
provided a number of models for inferring network structure from the reports of multiple, unreliable informants. An
underlying assumption of these models is that informants are explicitly probed for the relationships on which they
report, as with roster-based instruments such as Krackhardt's (1987) Cognitive Social Structure. When informant
information is obtained from unstructured sources such as personal testimony or organizational documents,
however, such probes are obviously impossible. Here, I demonstrate an extension of existing network inference
methods to the analysis of data from unstructured sources, incorporating model-based corrections for
non-ignorable sampling of edge reports. Applications of this approach to the analysis of police reports from
responders to the World Trade Center disaster and to interorganizational networks in the Hurricane Katrina
response are shown; implications for network inference from name generator data are also discussed.
Session: Formal Methods and Statistical Models
Keywords: network inference , statistical models , informant accuracy , unstructured data
Accepted: Yes

62 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Romina Cachia
Title: Social computing: Of users, networks and co-production
Inherent characteristics of online interaction are redefining the role of audiences, from passive participants in a
mediated process, to active agents catalysing basic aspects of mediated communication. The significance of this
evolution lies in the fact that online users are not only proactive, but have evolved into a collective networked
force. Participation in Online Social Networks sites (OSNs), such as MySpace, Flickr and Friendster is one facet
of this evolution.

In recent years, OSNs have gained unprecedented popularity, mainly through innovative provision of online
interactive platforms which enable free exchange of multi-modal content in a self-organised environment.
Building on existing work on collective action, this work attempts to illustrate the underlying factors determining
the transitory role of the users in social computing networks, so as to try and extract what prompts users to
produce, create and share content for free in online environments.

For many years, technology research overlooked the role of the user as a significant variable in studying
technology. This study proposes a triangulation of ego-centric network analysis and online in-depth interviews.
Extrapolating publicly available variables on OSNs, social networks of users' collaboration are explored to identify
usage trajectories according to the following variables: posting of content, text exchange, subsequent posting of
more content and changing affiliation amongst users' contacts along a temporal boundary. Flick, a photo sharing
application is used as a case study.

This study is a work-in-progress, as part of a doctoral programme

Session: On-Line Communities
Keywords: collective action , mixed methods , social networks , social computing
Accepted: Yes

63 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dragos Calitoiu
School of Computer Science, Carleton University, Ottawa, Canada and Health Policy Research Division, Health
Authors: Canada
Zachary Jacobson
Health Policy Research Division, Health Canada, Ottawa, Canada
Title: Dynamic Social Networks (dynamic weighted graphs) and the corresponding entropy-like measures
Current analyses do not provide measures of how people join, expand and leave social networks. However it is
clear that social networks are dynamic structures and their activities depend on time; actors move into networks
and out again and relations among the actors evolve. Network processes are series of events that create,
sustain, and dissolve social structure. To understand social networks more fully, we must study their dynamics
and evolution, introducing time and time-varying edges and network topologies in addition to traditional
descriptions of network structure.

A network’s evolution and temporal dynamic seem intuitively likely to modify the classical measures involved in
social network analysis [SNA]. Any cross-sectional description of a network, at any particular time, can only
describe a network state, not a process. A network event implies two discrete times and a description [perhaps
incomplete] of the state at those times. A process implies one or more events.

We propose entropy-like measures that can vary across time among network edges, and can provide a global
descriptor of the dynamic relations operating within a social network. We show preliminary results for the study of
social networks and draw inferences in the health domain.
Methodology: We add dynamic edge-weighting to classical network analyses:
A social network is weighted if each edge has an associated weight, w. If the weighting wij of the edge between
node i and j also varies as a function of time, wij(t), we say that the social network is dynamic. We adapt this
approach to classical measures for describing static social networks: degree centrality, closeness centrality, and
betweenness centrality.

The entropy of a network may be computed from the distribution of instantaneous edge weightings, E = S
P(X)log2P(X). The information content, or reduction of uncertainty in knowing the edge weights then is of course,
H = – E. Again, if E evolves in time as E =E(t), then the information capacity of the network must also change
with time.

Example: We show results in treatment and prevention of the spread of sexually transmitted diseases [STD’s]
and other infectious agents. We further apply the treatment to information communication. Clearly, there are
other important applications where this treatment will provide a richer description of network content and activity.
Session: Network Dynamics
Keywords: dynamics on networks , entropy-like measures
Accepted: Yes

64 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Penelope Canan
Authors: pcanan@mail.ucf.edu
University of Central Florida
Title: The Correspondence Between Policy and Implementation Networks
When policy "talks," implementation may not follow as the necessary actors may not be able to walk that talk.
Careful attendance to processes that support implementation by the right sectoral actors to accomplish
efficacious changes in performance--new instituional practices--is necessary. Lessons from the most successfully
implemented global environmental treaty in history, the Montreal Protocol on Substances that Deplete the
Stratospheric Ozone Layer, go beyond policy talk to policy tools that supported actual protection critical to
planetary survival
Session: Policy Networks on Climate Change
corporate social capital , collaboration , innovation , inter-organizational networks , interviews as data ,
network processes , relationship formation , individual performance
Accepted: Undetermined

65 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Geoffrey Canright
Authors: Telenor R&D, B6d, Snarøyveien 30, Fornebu 1331, Norway
Kenth Engø-Monsen
Telenor R&D, B6d, Snarøyveien 30, Fornebu 1331, Norway
Title: Epidemic spreading of electronic viruses on directed graphs
Spreading of electronic viruses, among computers and mobile phones, typically depends on address/phone
number lists. The network formed by these lists is not symmetric: the fact that A has B’s address does not ensure
that B has A’s address. Thus the underlying network on which such spreading takes place is directed: the links
are in general one-way. We present an extension of our “topographic” analysis for spreading on undirected
graphs, to the case of directed graphs. We find that some ideas from Web link analysis lead us to a concrete
prediction: that the epidemic coverage (after long time) changes qualitatively when the rate of infections from
“outside” the network exceeds a threshold rate. Specifically, for low rate of infections from outside, with high
probability, only the giant component and its out-components are infected; while for above-threshold infection
rate from outside, the whole graph is likely infected. Our theory gives an estimate of this threshold rate. Finally,
we present some limited tests of this theory using simulations on directed graphs.
Session: Infectious Diseases and Social Networks
Keywords: dynamics on networks , graph theory , model-based simulation
Accepted: Yes

66 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Julian Cardenas
Authors: University of Barcelona, Sociology, Av. Diagonal 690 Barcelona, Barcelona 08034, Spain
Josep A. Rodríguez
University of Barcelona, Sociology, Av. Diagonal 690 Barcelona, Barcelona 08034, Spain
Title: Expandind and conquering networks
In this paper we analyze and compare several takeover bids, seen as strategies used by networks of
corporations (what we will call mega-network corporations) to enlarge their social capital and power. Given the
Abstract: relevance of these mega-network corporations, takeover bids often fuel economic struggles and political actions.
Therefore, to understand the success of such takeover bids we need to look into network indicators as much as
into the struggle and political intervention. We analyze these takeover bids within global interlocking directorates.
Session: International Networks
economic networks , interlocking directors , international networks , merger and acquisitions , business
Accepted: Yes

67 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

William Carroll
Authors: wcarroll@uvic.ca
University of Victoria, Sociology Department, Victoria, British Columbia V8W 3P5, Canada
Title: Tracking the transnational capitalist class: the view from on high
This paper presents preliminary findings from a study of transitions in the global corporate elite between 1996
and 2005. It explores the extent to which and ways in which the network of interlocking directorates among the
world's largest corporations has been recomposed in recent years, as a function of structural developments such
as the rise of new centres of accumulation on the semi-periphery and the continuing economic integration of
Europe. On the basis of this view from on high, the paper discusses recent tendencies in transnational capitalist
class formation and their implications for global political economy.
Session: Politics and Interlocking Directorates
Keywords: business networks , governance , globalization , interlocking directors
Accepted: Yes

68 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Tiziana Casciaro
Authors: Harvard Business School Boston, USA
Miguel Sousa Lobo
Duke University Fuqua School of Business Durham, USA
Title: Interpersonal affect and task interaction: An analysis of causality
Does structure cause emotions, or do emotions cause structure? Psychologists and sociologists alike agree that
failures and successes in task interaction elicit affective reactions toward an interaction partner. What is less
clear is the extent to which the reverse is true: do affective reactions shape the choice of task partners? Is
interpersonal affect a bona fide determinant of task networks, or simply a by-product of them? Using data from a
longitudinal network study of a luxury goods retail company, we find that interpersonal affect predicts task
interaction consistently more strongly than task interaction predicts affect. Such manifestations of affective
primacy in the formation of work networks point to the relevance of affect in structural theories of organizational
Session: Network Dynamics
Keywords: emotions , intraorganizational networks , network dynamics , network structure
Accepted: Yes

69 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Susan Cassels
Authors: University of Washington, Center for Studies in Demography and Ecology, Seattle, WA USA
Steven Goodreau
University of Washington, Department of Anthropology, Seattle, WA USA
Title: Using Exponential Random Graph Modeling to Assess the Population-Level Impact of the Acute Stage of HIV
Virological research as well as modeling work suggest that the fraction of new HIV infections attributable to
source partners in the acute stage of infection may be high. Nevertheless, no study documenting the role of
recent infections on HIV transmission in a population has been conducted. Our paper uses a form of dynamic
exponential random graph modeling along with a stochastic simulation model of HIV transmission dynamics to
estimate the proportion of new infections attributable to recently infected source partners. Behavioral data are
taken from a study of sexual behavior in thirty communities across Zimbabwe, collected in 2003. The ERG
Abstract: models which we fit include momentary degree distributions, relational types (marital vs. non-marital) and
relationship duration, each of which is important for understanding transmission dynamics. The model is
implemented in a version of statnet – an R package for network estimation and simulation. Initial results suggest
that around 25% of new HIV infections are attributable to recent infections in the source partner; although this
effect may vary depending on the stage of the epidemic. Findings may be used to assess the implications for the
potential effectiveness of CTL vaccines and other interventions for reducing HIV transmission at the population
Session: Infectious Diseases and Social Networks
Keywords: hiv , simulation , egocentric network
Accepted: Yes

70 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Damon Centola
Authors: dcentola@rwj.harvard.edu
Institute for Quantitative Social Science Harvard University 1730 Cambridge St., S408 Cambridge, MA 02138
Title: Complex Contagions and Social Health
The theory of complex contagions argues that social networks that facilitate the spread of disease and
information may not be ideal for the spread of social behaviors. When adoption of a new behavior requires social
reinforcement, either to increase the behavior’s legitimacy (such as with a new norm), or to increase its credibility
(e.g., when adopting a new technology), multiple social contacts may be required before the behavior will be
adopted. The greater the risk of social exclusion, embarrassment, or personal cost associated with adopting the
behavior, the higher the “thresholds” for adoption are likely to be. Researchers have formally shown that unlike
disease and information, which only require a single contact to spread from person to person, “complex
contagions”, which require multiple sources of confirmation before adoption, may not benefit from the long ties
that give small worlds networks their increased reachability. Rather, “bigger worlds” networks, with greater local
clustering may actually be better suited to spread these behaviors. While these formal results have important
implications for a variety of public health applications (e.g., the diffusion of condom use, following-up on “at-risk”
screening, HIV prevention behavior), they still await empirical confirmation. We present an experimental design
to test the theory of complex contagions by studying the diffusion of health behavior in an empirical social
Session: social networks and health
Keywords: network structure , dynamics on networks , collective action , informal norms , small world
Accepted: Yes

71 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Maryse Chappin
Authors: m.chappin@geo.uu.nl
Universiteit Utrecht Department of Innovation and Environmental Sciences Utrecht The Netherlands
A dynamic perspective on the relation of successfulness of a collective knowledge generation project and its
network of participants
In this paper we address the relation of project outcomes of collective knowledge generation and network
configuration. We see collective knowledge generation as a learning process. During this learning process actors
originating from different organizations, are cooperating. This cooperation entails, among other things, the
exchange of information and knowledge among partners. The successfulness of the exchange is depending on
the structure of the network (e.g. closeness of actors) as well as on characteristics of participating actors (e.g.
level of commitment to the project). The question we ask is: In what way are learning processes and network
configuration of participants related over time?
Abstract: In this paper we have made a longitudinal database of the evolution of networks and learning processes. We
have developed this database out of archival data for a collective knowledge generation project of the Centre of
Competence Paper and Board, which consisted of four sub-programs. The nodes (the participants) are valued
with attribute data, such as presence during meetings, budget received from project or type of organization.
We expect to identify barriers and stimuli in terms of network configurations, which may explain why learning is
successful or not. Learning may, for instance, be hampered by frequently changing networks or expanding
networks due to a changing team composition. Whereas, the presence of brokers between the different
sub-projects, may contribute to a fruitful learning process.
Session: Inter-Organizational Networks
inter-organizational learning , network dynamics , social networks , structure of collaboration , knowledge
Accepted: Yes

72 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

David Chavalarias
Ecole Polytechnique, CREA, Paris, France
Jean-Philippe Cointet
Ecole Polytechnique, CREA & TSV (Social and Political Transformations related to Life Sciences and Life
Forms), INRA.
Title: Bottom-up scientific field detection for dynamical and hierarchical science mapping, methodology and case study
Modern acceptation of paradigm has been provided by T. Kuhn as "an entire constellation of beliefs, values and
techniques, and so on, shared by the members of a given community". He contended that, a paradigm enables a
group of scientists to focus its efforts on a well- defined range of problems. These paradigms are reflected at the
level of scientific production by paradigmatic fields, i.e. set of articles where there is a consensus concerning de
definition of important concepts and identification of problems and techniques needed to solve them. This
suggests that paradigmatic fields found in public sphere of knowledge production provide a direct insight into the
very structure of science and researchers communities: there is a deep correlation between the social community
structure and the conceptual community structure. A reconstruction of these paradigmatic fields is nowadays
possible. Massive collections of scientific publications are now available on-line thanks to multiple public
platforms. These databases usually cover large-scale scientific production over several decades and for a broad
range of thematic areas. The aim of this paper is to present tools for automatic bottom-up identification of
paradigmatic fields linked with scientific communities' structures from an article database. We define an
asymmetric paradigmatic proximity between concepts which provides hierarchical structure and test our methods
on a case study. The strength of our approach is that is does not require other information than the one already
available in most existing database to reconstruct the multi-scale structure of paradigmatic fields reflecting the
one of scientific communities.
Session: word networks or semantic networks
access to knowledge , paradigmatic evolution , publication analysis , construction hierarchy , history of
science , knowledge networks , kuhn , network evolution
Accepted: Yes

73 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

I-Chien Chen
Institute of Sociology, Academia Sinica
Authors: Chyi-In Wu
Institute of Sociology, Academia Sinica
Ming-Yi Chang
Institute of Sociology, Academia Sinica
A Best Friend Is Equal to A “Good” Friend? The Mutual Influence between Adolescent's Depressive Symptoms
and their Best friend's
Like other human relations, the friendship networks of adolescents are also always under dynamic transition.
Relations need to be established and maintained and, from the perspective of relation network, relations are
always under the processes of disrupting, changing, re-shaping and re-adapting. For adolescents, their
behavioral development is heavily influenced by their peer’s ideas and opinions and vice versa. Under the pattern
of interaction and within their own friendship network, adolescents exchange value, ideas, behavioral patterns,
emotions, and even attitudes. Thus, the stability of friendship network, just like best friend, should serve as a
critical basis of adolescent attitude and behavioral development.
According to this, the purpose of this study is to explore the influence of adolescents’ best friend and
adolescent’s mental and behavioral development. We focus our main concern on the mutual influence between
the development of adolescent depressive symptoms and their best friend’s. We argue that the best friend may
be not always equal to a “good” friend.
Abstract: In this study, we intend to pin down the data analyses on the parallel interaction between adolescent and his/her
best friends (across three wave’s observations). Further, using the Latent Growth Curve Model (LGC), we intend
to show the developing trajectories of both adolescent and his/her stable best friend to investigate the effect of
stable friendship on individual’s mental health.
The primary findings are, first, adolescents tend to make friends similar to themselves, i.e., happy kid looks for
cheerful kids to be friends and unhappy kids tend to be friend with kids with sad mood. However, there is no
significant evidence to show that there is any association between the change path of adolescent’s depressive
symptom and their best friend’s. Nonetheless, the further analysis of model comparison between adolescent who
have mutually recognized best friend and those who has not, indicated that the mutual influence of the
adolescent's depressive symptoms is accord with his mutually recognized best friend’s. These findings reflect
that for adolescent, a friend is a friend, but under some certain circumstances, a friend can become a
“meaningful” friend.
Session: Adolescent Friendship Networks
Keywords: adolescence , depression , friendship networks , reciprocity , longitudinal
Accepted: Yes

74 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Pauline Cheong
Authors: pcheong@buffalo.edu
University at Buffalo, Department of Communication, Buffalo, New York, USA
Missions on the Web, at home and abroad: (Re)examining Church life and the transnational connections of
Chinese Protestant Churches in Toronto
Social research has recently spotlighted how the Internet is facilitating changes in the social networks, norms,
and practices of community. In the wake of 9/11 and other unrests in several world cities, questions concerning
religious culture have come to the forefront of international attention and government intervention. This paper
discusses and examines the ways in which church life and transnational religious connections are built among
Chinese migrants by analyzing religious networks in two areas; organizational mission outreaches and
interpersonal discipleship.

Using network-mapping softwar-e, this paper performs an analysis of the websites of Chinese Protestant
churches in Toronto in order to investigate their missional relations and religious educational connections to othe
religious organizations in Toronto and abroad with other overseas Chinese communities, particularly in Asia. This
Abstract: quantitative analysis is supplemented by findings from focus group interviews held with Chinese Christians in
Toronto, which examines the ways and extent in which church leaders and laity are utilizing the Internet for
spiritual purposes.

Findings highlight how some religious organizations are increasingly utilizing web spaces to advance their
missions and enlarge ‘the body of Christ’. Hyperlink analysis provides understanding of the growth of
transnational religious communications; in this case among Chinese Christians who strategically construct
religious networks and circumvent censorship in ways sensitive to the info-political geographies and online
regulation of their host context, China. Interview data further address the significance and drawbacks of these
online connections for the building of ‘jia’, church family life for new migrant Chinese in contemporary society.
(248 Words)
Session: Communication Networks
social relationships , globalization , religious culture , civil society , cultural communities , ethnic identity ,
internet/hyperlink networks , migration networks
Accepted: Yes

75 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Chin-Shan Chi
Authors: Youth & Child Welfare Department, Providence University, Taichung, Taiwan.
Ray-May Hsung
Department of Sociology, Chengchi University, Taipei, Taiwan.
Title: The collaborative networks among long-term care organizations in Taipei city
This study attempts to analyze how the collaborative networks among small private long-term care organizations
are created under strong market competition and high uncertainty of this organizational field or industry. We are
also concerned about the factors on the creation of different types of collaborative networks. Furthermore, we
plan to explore the functions of these collaborative networks on the organizational effectiveness. The data are
from a survey of the long-term care organizations in Taipei in 2005. There were 187 long-term care organizations
in Taipei city, and we successfully complete 186 cases. There were 225 pairs of collaborative ties between
long-term care organizations. We found that the state evaluation pushed the density of collaborative networks
increased among these long-term care organizations. The density of imitation and learning networks is the
highest, the second highest is that of resident referral networks, and the third highest density is that of
collaborative purchase networks. On the dyadic level, the centrality of ego and alter is highly correlated. This
result implies that the long-term care organizations with high degree of centrality tend to cooperate with each
other. This might be explained by the status signal and homophily principles. On the individual level, the
organizations with high degree of centrality in the collaborative networks tend to has high organizational
Session: Inter-Organizational Networks
Keywords: centrality , effectiveness , homophily , inter-organizational networks , network structure , status transitions
Accepted: Yes

76 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Arul Chib
University of Southern California, Annenberg School of Communications. Los Angeles, California.
Authors: Aram Sinnreich
University of Southern California, Annenberg School of Communications. Los Angeles, California.
Jesse Gilbert
California Institute for the Arts, Los Angeles, California.
Title: Modeling Information Equality: A Real-World Networks Simulation
In this study, we build and test the second version of a stochastic, agent-based model of information diffusion,
called dFusion. The first version, presented at Sunbelt 2005, used three separate real-world data sets to
demonstrate a relationship between social and/or media latency and the information equality of a given network.
We define information equality as the extent to which members of a social network gain access to salient
information at relatively equivalent speeds, while latency refers to limitations in this diffusion process stemming
from characteristics of the social structure or media infrastructure. By focusing on differential, rather than
absolute, speed of access to information, we hope to create an evaluative framework for information technology
investment that accurately predicts the effects of such interventions on social equality.
dFusion version 2.0 incorporates recent developments regarding the optimal structure of real-world social
networks. We attempted to integrate rival perspectives on social network structure into a single, supportable
social network algorithm. This meant using a preferential attachment process, bounded by nodal aging and a cap
on nodal degree. Simulations run with dFusion 2.0 suggest that, while a diffusion model for scale-free networks
can be created, the resultant structure leads to almost instantaneous diffusion under a variety of situations. A
comparative network analysis of the two versions found differing network properties that may influence the
diffusion process.
These findings may have implications for the debate about the relevant social structure of real-world networks.
Examining whether simulated processes external to the social network replicate real-world processes may lead to
a better understanding of the validity of the underlying network structure.
Session: Simulation
Keywords: actor-driven modeling , model-based simulation , personal networks , scale-free network , simulation study
Accepted: Yes

77 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Catherine C H Chiu
Authors: catherine.chiu@cityu.edu.hk
City University of Hong Kong Department of Asian and International Studies Hong Kong China
Trust in Out-groups in Hong Kong: An Exploration of the Relationship between Structural, Action and Cultural
Dimensions of Social Capital
Social capital is a multi-dimensional concept. It has a structural, an action and a cultural dimension. Some
scholars combine two or three of these dimensions to construct a social capital index while others simply use
associational membership as a proxy for both the structural and cultural dimensions. In this paper, I disentangle
the three dimensions and explore their relationship. Using associational membership, voluntary services and trus
to represent the afore-mentioned three dimensions respectively, I show how variations in associational
membership and voluntary services influence trust in two out-groups in Hong Kong. The two out-groups are new
Abstract: immigrants from Mainland China and domestic workers from Southeast Asia. Discrimination against these two
out-groups is one of the major social problems in Hong Kong. The social capital school has proposed that
voluntary associations make their members more trusting. It is argued that membership of voluntary associations
not only creates in-group trust but also tends to be generalized to the society as a whole. Information on whether
or not associational membership influences attitudes such as tolerance for or trust in out-groups is scant. This
paper fills this gap. Data for analysis come from two surveys conducted among employed and non-employed
people in Hong Kong.
Session: Social Capital
Keywords: social capital , trust , associational membership , voluntary services , out-groups
Accepted: Yes

78 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dimitrios Christopoulos
Authors: dc.christopoulos@uwe.ac.uk
UWE-Bristol, Politics, UWE-Bristol, Frenchay Campus, Coldharbour Lane, Bristol, Avon BS16 1QY, UK
Title: Peer Esteem Snowballing: Sampling in surveys of experts
A consistent problem with key informant, elite and expert interviewing is the representativeness of sample
populations. Since studies that employ such techniques depend on a small number of respondents, they are
often classed as qualitative. The possibility of going beyond these classic approaches arises by employing
methods developed to explore hidden populations in network analysis. These would qualify as descriptive
Abstract: quantitative techniques since we cannot provide a robust measure of reliability. It is the case however,
particularly in the investigation of small populations of expert opinion, that we can be confident of surveying a
sizable proportion of that population. A case study of such a survey employing Peer Esteem Snowballing (PEST)
is offered in demonstration. The treatment of responses for outliers and determining the robustness of specific
questions is the subject of an appendix.
Session: Sampling methods
Keywords: snowballing , expert surveys , methodology , network surveys , informant surveys , method triangulation
Accepted: Yes

79 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sotiris Chtouris
Authors: University of the Aegean
Anastasia Zissi
University of the Aegean
Title: Social networks and the concept of social capital: Theoretical and methodological issues
Over the last decade, the concept of social capital has been widely linked with issues of economic, social and
health development. Despite its popularity, a number of considerations are raised concerning its theoretical,
methodological and ideological clarity. This paper aims at promoting the theoretical understanding of the social
Abstract: capital concept by reviewing the evidence on the linking with the theory of networks. Both contemporary
sociological perspective and social community psychology approach are used as operational frameworks. The
theoretical implications on the methodological development of indicators of social capital are discussed together
with the ‘dark side’ at an ideological level.
Session: Social Capital
Keywords: community of practice , family , social capital , social support , social networks
Accepted: Yes

80 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Vincent Chua
Authors: vincent.chua@utoronto.ca
University of Toronto, Department of Sociology, 725, Spadina Ave, Toronto, ON M5S 2J4, Canada
Title: The conditional effects of social networks on job earnings in highly meritocratic Singapore
While personal contacts may often be a useful means to labour market success, their positive effects need not
apply to all labour market situations in general. This paper investigates the role of social networks in a type of
labour market which places great emphases on educational qualifications and meritocratic recruitment,
Singapore. Using job seeking data from a representative sample of some 902 Singaporeans, I address the
question, “under what conditions” will “who you know” matter more or less relative to “what you know” and “for
whom”? My data suggests that personal contacts are not a commonly used resource among highly-educated job
seekers, especially for placements into public sector jobs. On the other hand, lower-educated job seekers are
Abstract: more likely to use personal contacts, but I argue that the effectiveness of their contact use is limited since they
hardly gain access to high-status contacts. This paper argues that even if cross-class contacts are useful for
lower-educated individuals, the overarching pattern of status homophily makes such boundary spanning ties
rarely available. The paradox is that while high-status social networks may be beneficial for lower-educated
groups, segregative patterns which are probably the by-product of talent selection and early streaming in schools
prevent these groups from experiencing new arenas of change and resources. Finally, the relative unimportance
of job seeking ties for highly educated individuals is explained in terms of Singapore being a ‘coordinated labour
economy’ whereby education and training systems are tightly coupled with the labour market.
Session: Social Capital
Keywords: contact use , social capital , meritocracy
Accepted: Yes

81 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Louise Clark
Imperial College London, Centre for Environmental Policy, Edificio Hermanos Gonzales Depto 3 Piso 4, Calle Dr
Luis Uria de la Oliva #2661, La Paz, Bolivia
Title: Understanding Information and Innovation in Rural Bolivia: How can Social Network Analysis help?
Social Network Analysis (SNA) can help researchers and development practitioners to understand the complex
network of inter-institutional relationships implicit in the rural sector by providing tools with which to visualise
information flows and identify key actors influencing innovation amongst smallholder farmers. This paper
presents key methodological steps to using SNA as a diagnostic tool, with advice on designing surveys, cleaning
databases and creating network maps based upon experience of field work in Bolivia.

Surveys should be short and simple with clear questions and easily codified answers that define the qualitative
values that differentiate social groups as well as the relationships that exist between them. Data is codified in
excel spreadsheets and input directly to Netdraw, to make the methodology more easily accessible to Southern
development practitioners. These network maps create an x-ray of inter-institutional relationships in a given area
while partial maps show information flows related to specific topics or between certain social groups, helping to
identify general tendencies. Examples of 1-mode maps demonstrate the complexity of information flows in
agricultural supply chains while 2-mode networks show the relationship between actors and innovations.

Maps are discussed with the actors identified who draw their own conclusions, despite having no previous
experience of SNA. This approach can help researchers gain trust with a mechanism to return the data collected
to the subjects of the investigation, while enabling one to speak with openness and authority on the delicate
subject of inter-institutional relationships.
Session: Qualitative Network Research: Data Collection
information content , innovation , intraorganizational networks , network surveys , supply chain ,
visualization , access to knowledge
Accepted: Yes

82 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jean-Philippe Cointet
CREA, CNRS/Ecole Polytechnique, Paris, France & TSV, INRA, Ivry sur Seine, France
Authors: Carla Taramasco
CREA, CNRS/Ecole Polytechnique, Paris, France
Camille Roth
Dept of Sociology, Univ of Surrey, UK & CREA, CNRS/Ecole Polytechnique, France
Title: Socially-mediated concept diffusion in a scientific community
Knowledge diffusion is now widely assumed to be socially mediated, while it is unclear to what extent some
particular type of social interactions may indeed account for the propagation of a given concept or, put differently,
in which manner a concept may diffuse through a given kind of social network. To this end, we aim here at
empirically measuring concepts transmission through collaborations in a scientific community. The collaboration
network is appraised as a diffusion tree on which concept transmission occurs. More precisely, we observe
concept propagation along temporal series of collaborations and focus on authors who start using a given
Abstract: concept. We measure the likeliness for an author to have started using the concept owing to a collaboration with
at least one other author who previously used this concept. We compare this likeliness to the expected likeliness
under a null hypothesis where such collaboration occurs randomly. In other words, we measure the extent to
which this adoption may be driven by social factors, where "social" means in our case "exclusively
collaboration-based". The empirical study is based on a scientific collaboration network extracted from a
database including all publications concerning a precise, well-bounded field, the study of the "zebrafish" among
embryologists over the last twenty years.
Session: Knowledge Networks
Keywords: social influence , scientists , paper co-atuhorship network , epidemiology , dynamics on networks , diffusion
Accepted: Yes

83 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jordi Colomer Matutanos

Facultat de Informatica de Barcelona Universitat Politecnica de Catalunya Barcelona, Spain
Authors: Mariano Belinky
Departament de Llenguatges i Sistemes Informàtics Universitat Politècnica de Catalunya Barcelona, Spain
Bruce Kogut
Professor of Strategy Insead Business School Fontainebleau, France
Title: Visualizing Network Dynamics
In this paper, we present a novel technique to visualize datasets containing spatial and temporal network flow
information. We work with investment transactions between venture capital firms and start-up companies in the
US during a 45-year period. The visualizations show investment sources and targets geographically and over
time. The visualization permits an intuitive analysis of the geographic growth of the venture capital market and
suggests that contrary to static statistical analysis, local regions developed in response to across regional
investment flows. Our aim is to discover relevant feaures in the evolutionary dynamics of the networks, which
Abstract: tend to be hard to spot using traditional analytical techniques.
We implement and extend several network visualization layouts adding the ability to highlight both spatial and
temporal regions which show strong activity; we call this 'hot spots analysis'. We then use a cartogram approach
to overlap and correlate secondary economic information which can be easily contrasted to the network-related
phenomena. With this approach our visualization is able to help domain researchers answer questions regarding
the relationship between network dynamics and economic factors. With this mixed approach, we can use our
visualization tool to focus on particular features which are then analyzed using classic methods.
Session: Visualization
visualization , venture capital , affiliation networks , business networks , network evolution , networks across
Accepted: Yes

84 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Noshir Contractor
Science of Networks in Communities Research Group, National Center for Supercomputing Applications,
University of Illinois at Urbana-Champaign, 1205 West Clark Street, Room 1008, Urbana, IL 61801, USA
Harold D. Green Jr., York Yao
Science of Networks in Communities Research Group, National Center for Supercomputing Applications,
Authors: University of Illinois at Urbana-Champaign, 1205 West Clark Street, Room 1008, Urbana, IL 61801, USA
Andy Don, Chunke Su
Science of Networks in Communities Research Group, National Center for Supercomputing Applications,
University of Illinois at Urbana-Champaign, 1205 West Clark Street, Room 1008, Urbana, IL 61801, USA
Steven Harper, Nat Bulkley, Alex Yahja
Science of Networks in Communities Research Group, National Center for Supercomputing Applications,
University of Illinois at Urbana-Champaign, 1205 West Clark Street, Room 1008, Urbana, IL 61801, USA
Title: CI-KNOW: Advancing the Theorizing and Implementation of Multidimensional Network Recommender System
As communication and collaboration technologies become more pervasive in the emerging cyberinfrastructure,
there is a growing interest in leveraging these technologies to develop more effective and efficient referral
systems that provide members of the community with recommendations for appropriate resources (people,
documents, datasets, analytic tools, algorithms, etc.). In response to needs expressed by science and
engineering communities using cyberinfrastructure, the Science of Networks in Communities (SONIC) research
group at the National Center for Supercomputing Applications (NCSA) has been developing a network-based
recommendation system called CI-KNOW (CyberInfrastructure Knowledge Networks On the Web). We
conceptualize knowledge networks as intrinsically multi-dimensional, created from the interactions and
interconnections among the scientists, documents, data, analytic tools, sensors, etc.).

While there has been considerable efforts to develop recommender systems over the past decade, CI-KNOW
addresses three limitations of existing tools: First, in making a recommendation, CI-KNOW takes into account the
attributes of the resource being recommended as well as the network ties connecting that resource to the person
requesting the recommendation. Second, the algorithms used to make these recommendations take into account
computational advances in data mining as well as insights based on social motivations for information retrieval as
outlined in the multi-theoretical multilevel (MTML) model. Third, the data used to make these recommendations
are harvested from (i) individuals’ digital traces from the Web and bibliographic data sources, (ii) individuals’
actions, interactions and transactions within the cyberinfrastructure portal, and (iii) user-generated content on the
cyberinfrastructure portal such as tagging and social-bookmarking of resources using Web 2.0 technologies.

In our presentation we demonstrate the key theoretical and technological aspects of the CI-KNOW system. We
outline the theoretical mechanisms that are encoded into the network recommendation algorithms. We
demonstrate the use of interactive network maps that enable users to identify recommendations of resources
(people, documents, tools, data, etc.) and explore connections within the knowledge network. The development
of tools such as CI-KNOW provides an unprecedented opportunity for researchers to theorize and empirically
test the dynamics of large scale knowledge networks. It also creates opportunities for collaboration between
social network theorists and computer scientists to develop algorithms that make theoretically grounded
recommendation based on the structure and dynamics of the community’s network.
Session: Knowledge Networks
access to knowledge , collaboration , data mining , information content , interdisciplinarity , knowledge
networks , multi-level networks , mtml model
Accepted: Yes

85 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sarah Cope
Authors: Sarah.Cope@homeoffice.gsi.gov.uk
Home Office, London, UK
Using Social Network Analysis to Study the Behaviour of an Organised Crime Group and its Response to Police
The UK’s Home Office has responsibility for cutting crime, especially violent and drug related crime. To do this
effectively it requires robust evidence about the behaviour of organised crime groups involved in the supply of
drugs and how these groups react to police interventions. Recently the Home Office conducted a study to
determine whether SNA can be used to provide this evidence. An organised crime group involved in the supply o
drugs in an area of the UK was studied retrospectively. The data used in the study were taken from police
Abstract: intelligence reports spanning a 3 year period. The analysis showed that SNA was able to provide useful
evidence, such as: the group were operating in a hierarchical structure; the position a network member held in
the hierarchy was influenced by their role in the network; the structure of the social network was different to the
task network; how the structure of the network changed in response to police interventions; and that the group
were still able to operate even after the arrest and subsequent incarceration of the group leader. This paper will
discuss how these results and others were derived.
Session: Criminals, Gangs, Terrorists, and Networks
Keywords: organized crime networks
Accepted: Yes

86 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Rense Corten
Utrecht University, Sociology, Heidelberglaan 2, Utrecht 3584 CS, The Netherlands
Authors: Michal Bojanowski
Utrecht University, Sociology, Heidelberglaan 2, Utrecht 3584 CS, The Netherlands
Bastian Westbrock
Utrecht School of Economics Utrecht University Janskerkhof 12 3512 BL Utrecht The Netherlands
Title: The Dynamics of the Global Network of Inter-Firm Alliances
This paper studies the dynamics of the network of inter-firm R&D alliances on an
international and cross-industry scale, over the period 1990 - 2002. Previous research has
shown that the number of collaborative relations between firms has substantially
increased. A drawback of many of those studies is however that they focus only on the
firms that participate in at least one alliance, thereby neglecting changes in the total
number of firms. We re-examine empirical regularities from previous studies on R&D
alliance networks from an explicit network analysis perspective by complementing data on
established alliances with data on firm populations. We study the network evolution in
terms of overall density as well as regional and sectoral segregation using an extended
Abstract: version of Freeman's segregation index. Moreover, we investigate the development of
governance modes for R&D partnerships.
In the first place we find that, controlled for the changes in the firm population, the
network of R&D alliances is very sparse and becomes even less dense over time. Second,
segregation by nation is decreasing, indicating globalization. Third, segregation by
industry is consistently higher than segregation by nation, but shows no clear temporal
pattern. Fourth, the share of joint ventures among R&D alliances is decreasing in favor of
more flexible governance modes, suggesting an increase in trust. Finally, neglecting
changes in population sizes introduces significant biases to segregation analyzes but does
not seem to influence general descriptive statistics.
Session: Inter-Organizational Networks
economic networks , globalization , inter-organizational , methodology , network dynamics , network
evolution , regional clusters , segregation
Accepted: Yes

87 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Rochelle Cote
Authors: University of Toronto, Dept. of Sociology, 725 Spadina Ave., Toronto, ON M5S 2J4, Canada
Bonnie Erickson
University of Toronto, Dept. of Sociology, 725 Spadina Ave., Toronto, ON M5S 2J4, Canada
Title: Diverse Ties, Diverse Effects: Looking to Networks to Help Explain Tolerance towards Ethnic Minorities
What is the role of social networks in generating positive orientation towards minorities? Past work includes
debates concerning both which network variables encourage tolerance, and how network variables connect with
other sources of tolerance. These debates have been inconclusive
because earlier research did not include all of the necessary variables in the same data set. We report results for
the first survey to be so inclusive, the 2004 Canadian federal election survey. Our measure of tolerance is a
highly reliable scale of positive attitudes towards ethnic and racial minorities and immigrants. We find strength in
Abstract: weak ties: tolerance is not related to the racial diversity of friends, but does increase with the variety of middle
class contacts and decrease with the variety of working class contacts (measured through a position generator).
Different kinds of sources of tolerance have independent effects, contrary to some earlier arguments that network
effects may be spurious with respect to important individual attributes such as education and occupational
prestige, and that the effects of attributes and voluntary association participation may work primarily through their
impacts on networks. The pattern of results is consistent with both the contact hypothesis and the competition
Session: N/A
Keywords: social networks , ethnic minorities , tolerance , canada
Accepted: Yes

88 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sarah Cotterill
Authors: busshc@leeds.ac.uk
University of Leeds England
Title: Using Social Network Analysis and qualitative research in case studies of public sector partnerships.
An increased role for electronic government is a central plank of the UK government’s modernisation agenda,
with the aim of improving local government efficiency and increasing its accessibility and responsiveness to local
citizens. This research examines the network of relationships between individuals and organisations involved in
sub-regional e-government partnerships in the UK. These partnerships variously involve officers from different
departments within a number of local authorities, councillors and other public organisations such as police, fire
and health services. The research explores how local authorities and their partners work together to implement
electronic government, identifying the issues, challenges and successes.The policy contribution will evolve from
Abstract: an increased understanding of the social networks underpinning complex service innovation.
This paper focuses on the methodology adopted during the study: a systematic literature review and comparative
case studies of three sub-regional e-government partnerships using a combination of social network analysis and
qualitative interviews. Network data was collected from participants using a short questionnaire to ascertain who
they dealt with in relation to e-government. This data was analysed using Ucinet. The social network data was
then used during qualitative interviews and workshops to generate discussions on why the network looks the way
it does and learn more about the meaning behind the surface of the relationships. This approach has helped
create a richer picture of e-government partnerships than could be found from social network analysis alone.
Session: Mixed Methods: Combining Qualitative and Quantitative Approaches
Keywords: social networks , mixed methods , local government , governance , partnership
Accepted: Yes

89 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Bruce Cronin
Authors: b.cronin@greenwich.ac.uk
University of Greenwich, Business School, Park Row, Greenwich, London, UK SE10 9LS, UK
Title: The Decline of the New Zealand Business Roundtable
New Zealand’s experiment with radical neoliberalism in the 1980s was closely associated with the establishment
of the country’s most vociferous lobby group, the New Zealand Business Roundtable. From its origins as a
reaction by Auckland manufacturers to a surge of industrial action in the mid 1970s to the conscious imitation of
its US namesake and formalisation from 1985, the group’s claims to speak for New Zealand’s biggest businesses
have been widely accepted. Yet a closer examination of the place of this lobby group in the nexus of power
relations among large New Zealand firms raises major questions about the role of this group in the country’s
politics and the neoliberal policies it championed.
This paper traces the evolution of the New Zealand Business Roundtable in terms of the network of social
connections forged by interlocking directorships from its origins in the 1970s, through its political ascendancy in
the 1980s and its decline thereafter. It is argued that the group represented a particular section of the business
community, the financial sector, rather than business, or even big business, in general. It is further argued that
the lobby gained its influence from the relative power of the finance sector in business in the 1980s, more than
any autonomous activity, and that the lobby’s influence declined with the marginalisation of this sector in the
Session: Politics and Interlocking Directorships
Keywords: political networks , interlocking directors , social capital , informal networks , policy network
Accepted: Yes

90 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Juan Cruz
Authors: Intelligent Systems Research Laboratory, National University of Colombia, Bogotá, Colombia
Fabio González
Intelligent Systems Research Laboratory, National University of Colombia, Bogotá, Colombia
Title: A Social Network Based Model for e-Mail Information Visualization
This paper presents a model for visualizing the information contained in a set of personal e-mails. E-mail
messages have a large amount of information; such information, in general, is not always visible to its owners.
For instance, e-mail users are usually aware of their contacts, but they are not necessarily aware of the
relationships between them. This information becomes explicit when a contact network is built based on the
information found in the header of a set of e-mails. In the same way, the set of topics, which reflect the interests
of the user and her/his contacts, could be made explicit by a topic map. This topic map is built based on the
keywords found in the e-mail subject, body and attachments.

The paper proposes an approach for visualizing the contact network and the topic map in an integrated image.
First, the topics or categories are identified by applying a clustering algorithm to a feature vectores extracted from
the e-mail, body, subject and attachments; second, the contact network is built based on the from, to, cc and bcc
fields of the e-mails in the set; third, the contact network is visualized as a social network, using a new layout
algorithm, and the network topics categorization information. The layout algorithm is based on an egocentric
configuration of the network: the main user is drawn at the center of the network, the contacts are drawn in
different concentric circles around the central node depending on the degree of separation from the main user.

Topics are displayed as "clouds'' surrounding some group of nodes. Those clouds may overlap since different
people may handle intersecting sets of topics.

The model was implemented in Java, and was tested using the Enron data set.
Session: Visualization
Keywords: information visualization , egocentric network , graph layout
Accepted: Yes

91 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Alexandra Curley
Authors: acurley22@gmail.com
Northeastern University, Center for Urban and Regional Policy, Boston, MA, USA
Title: Dispersing the poor: Relocation effects on social networks
This paper reports the results of an ethnographic study that followed residents from a Boston HOPE VI
redevelopment site throughout the relocation period. Findings focus on the effects of relocation on residents’
social networks—an important indicator of social capital. The goal of the study was to examine the dynamic
changes in low-income peoples’ social networks as they relocate to different types of communities. The research
examined the effects of HOPE VI relocation on residents’ social networks and found that residents had ties to
Abstract: people who provided social support and social leverage, as well as people who were draining. Overall, relocation
had a negative impact on participants’ supportive social ties; a positive impact by reducing draining ties; and no
impact on improving access to social leverage through bridging ties or the formation of new ties. Some women in
the study experienced a combination of positive and negative changes in their social networks. The findings
suggest that relocation effects on social networks can vary from person to person, and are not uniform the way
some theoretical perspectives and policymakers might expect.
Session: Qualitative Network Studies - II
qualitative research , changes in networks , exchange networks , functioning of social capital , friendship
networks , job seeking , negative ties , neighborhood community
Accepted: Yes

92 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Peter Dahlin
Authors: peter.dahlin@mdh.se
Malardalen university, School of Business, PO-Box 883, Vasteras S-72123, Sweden
Title: Joining the Forces of SNA and the European Business Network Perspective
Since the late 1970ies research has been made on business networks as a way to approach companies’
interaction and acknowledge the mutuality in the business-to-business setting. This interaction contains
exchange of various kinds, and behavioural aspects such as trust and commitment are used to describe what
constitutes these long-term business relationships. The influence from sociology is thus evident. The business
relationships affect each other, and this connectedness is what makes up the reason to describe a set of
business relationships as a business network. Studying business networks empirically is hard. Qualitative case
studies with focus on the parts, i.e. the actor and relationship, dominate. A lack of quantitative studies and focus
on structural aspects of the business networks have been noted, which opens up for making use of methods and
measures from the vast field of SNA. If using the methods, measures and reasoning with caution, in respect to
the differences, there are great opportunities for taking the business network concept further by learning from the
field of social network analysis.
This paper describes how SNA can be applied to business networks, and issues that must be handled. Previous
attempts to borrow methods and reasoning from sociology have ended unsuccessfully due to the inexpedience of
transferring some assumptions. For example, a person is able to have relationships with every other, whereas a
company can only do business with companies supplying or demanding their resources. Furthermore,
delimitation issues are described and studies of change examplifies a topic which could benefit from a structural
The paper describes an actual set of data in which SNA is applied on business networks, and the discussed
issues are described from the actual data set. The data has been collected from newspaper articles and
describes mergers and acquisitions among Swedish IT-related companies during 1994-2003, holding more than
3000 companies, over 1300 M&As and over 3100 relationships.
Increased understanding of the structural mechanisms of business networks could take the business network
field further. By approaching business networks as structures, and reaching the characteristics and mechanisms
of it, SNA could make the business network concept more than just a metaphor!
Session: Business Networks
Keywords: business networks , dynamics on networks , inter-organizational , network structure
Accepted: Yes

93 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

James Danowski
University of Illinois at Chicago, Communication, MC 132, University of Illinois at Chicago, Chicago, IL 60607,
Title: A Comparison of Software for Automated Network Analysis of Text
Since network analysis of text, based on coding word cooccurrences, was introduced (Danowski, 1982; 1986;
1988) several software packages for automated statistical analysis of text have been reported in the literature:
WORDLINK (Danowski, 1986; 1988; 1993a; 1993b, 2007), CATPAC (Woeflel, et al., 1992), VBPRO (Miller,
1997), CRAWDAD (Corman, 2002), AUTOMAP (Diesner & Carley, 2004, and LEXIMANCER (Smith &
Humphreys, 2006). Each of these packages enables some form of network analysis of text, either through built-in
routines or by formatting word cooccurrence data for input to standard network analysis packages such as
While most of these network content analysis programs have a somewhat similar approach to identifying word
cooccurrences, the programs have different pre-processes, post-processes, graphics, input file limits, data
processing capacities, and data import/export features. To date, none of the programs except WORDLINK and
CATPAK have been compared using the same textual corpora, which lead CATPAK to adopt the WORDLINK
algorithm for identifying word cooccurrences.
This paper reports on a comparison of the six word cooccurrence-based packages using a large textual corpora
consisting of the full text of all email exchanged in an organization over an 18-month period. The relative
advantages and disadvantages of the software packages are discussed based on the empirical results.
Recommendations are made for the most effective applications of the packages in analyzing text networks.
Session: word networks or semantic networks
automated text network analysi , word networks , semantic networks , network-based content analysis ,
network analysis of text
Accepted: Yes

94 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Galina Daraganova
University of Melbourne, School of Behavioural Science, Department of Psychology, Victoria, Australia
Pip Pattison
Authors: University of Melbourne, School of Behavioural Science, Department of Psychology, Victoria, Australia
Garry Robins
University of Melbourne, School of Behavioural Science, Department of Psychology, Victoria, Australia
Peng Wang
University of Melbourne, School of Behavioural Science, Department of Psychology, Victoria, Australia
Title: Social influence models
Social influence models
Galina Daraganova, Pip Pattison,
Garry Robins, Peng Wang
University of Melbourne,
School of Behavioural Science,
Department of Psychology,
Victoria, Australia
Abstract: In this paper we review a general class of models for location-dependent interactive social processes originally
developed by Robins et al (2001). Here we present an extension of the social influence model based on
non-Markovian neighbourhood assumptions. The model expresses interdependent actor attributes as a function
of exogenous relational variables, other exogenous attribute variables, and spatial location. Model parameters
reflect a variety of different influence effects, and we discuss the theoretical basis of each of these effects.
The application of the model is discussed in the context of a study designed to examine the potential role of
social networks and individual level factors in understanding persistent patterns of spatial clustering in
unemployment in Australia.
Session: N/A
Keywords: social influence , social network , network structure , attributes
Accepted: Yes

95 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dietrich Darr
Technische Universität Dresden, Institute of International Forestry and Forest Products, Pienner Str. 7, D-01737
Authors: Tharandt, Germany
Jürgen Pretzsch
Technische Universität Dresden, Institute of International Forestry and Forest Products, Pienner Str. 7, D-01737
Tharandt, Germany
The influence of structural and functional network properties on the spread of agroforestry innovations within
farmers groups – Evidence from Eastern Africa
Novel ideas and farming practices spread, and often enough even evolve, through interpersonal interaction and
communication in rural communities. Hence, facilitating cooperation and exchange among farmers becomes a
core objective of most extension interventions. Formal and informal farmers groups can play an active role to
promote the diffusion of knowledge and technologies and thus to improve the efficiency and efficacy of the
technology extension efforts.

This paper explores structural and functional group-level and ego network-level characteristics of formal and
informal farmers groups to explain the spread of agroforestry innovations and tree management related
knowledge among the group members. Specifically, it was hypothesised that at the group network level group
activity, composition and cohesiveness facilitate the spread of the technologies among group members; and
household centrality and prestige are positively related to household-level innovativeness. The technology
adoption behaviour of individual farm households serves to operationalize the diffusion variable.

Four full sample surveys of approximately 200 households each have been conducted in two districts of Kenya
and Ethiopia, respectively. Primary data was collected using semi-structured questionnaires, expert interviews,
group discussions and rankings, as well as observation. Data was analysed employing sociometric and statistical
software packages.

The research results clearly support the main hypotheses. Recommendations refer to the improvement of formal
and informal farmers groups in order to facilitate the agroforestry extension work.
Session: Innovation
Keywords: diffusion , forestry , farmers groups
Accepted: Yes

96 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Surja Datta
Authors: surja.datta@gmail.com
University of the West of England, School of Strategy and International Business, Bristol, UK
Title: It's a Small World: Social Identity and Trust in Offshore Outsourcing
This paper argues that social ties that are made on a substrate of social identities are a source of the initial trust
that is required for procuring services from geographically dispersed regions. The study is based on affiliation
networks of the founder entrepreneur of 40 successful outsourcing firms based in India. The study finds that a
significant number of these entrepreneurs had major North America affiliations prior to setting up their firms in
India, a fact that assumes significance in a context in which over 60% of offshore outsourcing transactions takes
place out of North America. Integrating social network with transaction cost economics also provides a better
Abstract: conceptual framework to explain the growth of offshore outsourcing that has traditionally been explained in terms
of labour cost arbitrage. Looking at offshore outsourcing as transactions that are embedded in social ties can
explain why such transactions have been limited only to certain geographic regions. In the presence of very few
technological barriers, the question to be asked is not why such growth, though exponential, has been witnessed
but rather why the growth has not even been higher. Conceptualising social ties as a resource that fosters initial
trust building amongst network partners leads to the suggestion that there are inherent limitations to growth of
such transactions.
Session: Social Capital
Keywords: cognitive networks , social identity , inter-organizational , social capital , trust
Accepted: Yes

97 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

George Davis
Carnegie Mellon University, CASOS Lab / COS Program / School of Computer Science, 244 39th St., Pittsburgh,
PA 15201, USA
Title: Networks from Paths: A Relational Perspective on Geospatial Traces
Network analysis is the study of structure in simultaneously existing relationships. However, these relationships
can often be directly observed only through a series of short lived and mutually exclusive interactions over time.
As an example, this talk examines two datasets which track the sequential movements of merchant vessels from
one port to another within a region. We present several methods for extracting network data from these
sequential observations, ranging from simple algebraic methods to probabilistic graphical models which combine
a variety of evidence to infer possible relations influencing movement patterns. We explore the advantages and
disadvantages of each technique, and briefly discuss the reverse problem: generation of sequential behaviors
from networks.
Session: Methods and Statistics
Keywords: spatial structure , 2-mode , bipartite networks , statistical models
Accepted: Yes

98 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Kristin Davis
International Food Policy Research Institute PO Box 5689 Addis Ababa, Ethiopia Tel: +251.11.645-8812 (From
the US dial 1-650-833-6696 x 2506) Fax: +251.11.646-2927
David Spielman
International Food Policy Research Institute PO Box 5689 Addis Ababa, Ethiopia
Martha Negash
International Food Policy Research Institute PO Box 5689 Addis Ababa, Ethiopia
Title: Links to innovation: an analysis of ethiopian smallholder networks
This paper describes how social networks contribute to the ability of Ethiopian smallholders to innovate, that is, to
make use of new or existing knowledge and technology in their decisions about agricultural production. Social
network and institutional ranking analyses are used to illustrate how networks contribute to smallholder
innovation, based on data gathered from participatory rural appraisals and supplemented by household-level
survey data, document analysis, and key informant interviews.
Findings demonstrate important differences between smallholders in a given community and between
communities, including network size, and sources of knowledge/information, inputs/materials, credit/finance, and
market linkages/price information. We conclude that public sector services play an important role in enabling
smallholder innovation, and that alternative service providers are of less importance. Given the persistence of
rural poverty combined with a drive toward greater smallholder commercialization, these findings suggest the
need for further diversification in the role of both market and non-market actors to foster greater innovation.
Session: Social Capital
Keywords: access to knowledge , methodology , social capital
Accepted: Yes

99 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ainhoa de Federico
Maître de Conférences, Institut de Sociologie et d'Anthropologie, Université de Lille 1, IFRESI, 2 rue des
Canonniers, 59000 Lille, France.
Title: Cross-national friendships and European identifications
Macro belongings to abstract "imagined communities" (Anderson 1983), such as nations, are grounded on the
generalized inductive transfer, encouraged by institutions of the interpersonal experience of trust and solidarity.
Trust and solidarity are first learned in the interactions taking place sphere of kinship, then in friendship that goes
beyond primary family solidarity into more abstract solidarity in universalistic societies (Eisenstadt 1956).

The national social contract is founded on the ideal model of friendship solidarity. The ideal fraternity announced
by the nation serves to hide and justify , the power and instrumentality that the order of the nation-state exerts
(Eisenstadt 1984). Meanwhile, friendship sustains the social contract while it is relegated to the private sphere
due to its subversive potential (Paine 1969).
What happens then with cross-national friendships? The paradoxical figure of the "foreign friend" introduces
tensions between the foundations of abstract solidarity and the experience of interpersonal solidarity.
According to the institutional logic of national states a cross-national friend should be considered as a foreigner,
but to the logic of friendship and interpersonal trust and solidarity, a friend almost by definition could not be
considered either stranger nor foreign. Much less if the belonging of this cross-national alter is guaranteed by his
embeddedness in a network of trustworthy relations.

Using hierarchical regression analysis we examine the effects of macro (institutional and ideological) and meso
(dyadic friendship and trust guaranteed by the network) factors on the categorization of cross-national friends in a
sample of 218 European exchange students.
Session: Networks and Identifications
Keywords: cross-national friendship , identification , friendship networks , european identity
Accepted: Yes

100 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Tjip De Jong
University of Twente, Faculty of Behavioural Sciences, tdejong@kessels-smit.nl, Enschede, Overijssel 3581CD,
The Netherlands
Title: Social capital, social networks and knowledge productivity
This paper presents a theoretical framework on the relation between social capital, social networks and
knowledge productivity within organizations. In our knowledge economy, the competitive advantage of
organisations relies on the capability to adapt to the changing environment by the continuous generation and
application of new knowledge (Harrison & Kessels, 2004). This specific capability is described as knowledge
productivity. In this paper and the related study the construct of knowledge is described as a social process of
interaction between organizational members. If we consider organizational knowledge as a ‘social process of
knowing’ (Huemer, Von Krog & Roos, 1998. p.140), social capital and the supportive theories offer a conceptual
framework for study and understanding this process. This perspective provides argumentation that the
composition and relation between social capital and social networks affects the organizational ability to become
knowledge productive. This paper will theoretically conceptualise that knowledge productivity can be described
as a continuous learning process. Henceforth, this paper elaborates on the notion that these learning processes
can only take place in a social environment in which the structural and relational perspective of social networks
play a facilitating role. This paper explores the dynamics of these powerful learning processes within social
networks that take place in a rich landscape of social capital.
Session: Innovation
Keywords: intellectual capital , social capital , access to knowledge , innovation
Accepted: Yes

101 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Maurits De Klepper
VU University Amsterdam, Public Administration & Organization Science, De Boelelaan 1081, Amsterdam,
Noord-Holland 1081HV, Netherlands
Gerhard Van de Bunt
VU University Amsterdam, Social Science Research Methodology, De Boelelaan 1081, Amsterdam,
Noord-Holland 1081HV, Netherlands
Peter Groenewegen
VU University Amsterdam, Public Administration & Organization Science, De Boelelaan 1081, Amsterdam,
Noord-Holland 1081HV, Netherlands
Title: The dynamics of intra-organizational advice networks
This paper addresses the question of how advice networks evolve over time within organizations. A range of
papers argue (Ibarra, 1992; Casciaro & Lobo, 2005; Nebus, 2006) that individuals initiate and maintain work
relationships such as advice because of affective and instrumental motives. Empirical studies investigated these
motives simultaneously using exogenous attributes such as gender similarity and skill complementarity (e.g.
Hinds et al., 2000). Few studies, however, examined these two motives together with (endogenous) network
mechanisms - such as the tendency to establish reciprocal and transitive relationships - despite that in
intra-organizational network studies these mechanisms have been shown to be relevant (e.g. Van de Bunt,
In our paper we combine affective and instrumental motives with these network-mechanisms to test a more
comprehensive model. Specifically, we aim at an examination of advice-asking by looking at the temporal
opportunity structure within the organization, task dependencies between co-workers, similarity in work values,
friendship between co-workers, complementarily in task skills and network effects like reciprocity and subgroup
We present the results of a three-wave longitudinal network study that took place at a Dutch hospital department.
The data are analyzed using a simulation-based statistical inference method for the evolution of social networks,
called SIENA.
Session: Intra-Organizational Networks
Keywords: network dynamics , intraorganizational networks , informal networks , actor-driven modeling
Accepted: Yes

102 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Wouter De Nooy
Erasmus University Rotterdam, History & Culture Dept. (L2-12), P.O. Box 1738, Rotterdam NL 3000 DR, The
Title: Matthew landscapes. Segmentation and concentration in reward systems visualized.
In his seminal paper on reward systems in science, Robert K. Merton introduced the Matthew Effect (Science
159 (1968) 5, 56-63), viz., the tendency of rewards to accumulate. Merton showed that the awarding of a Nobel
Prize to scientists increased the probability of receiving other types of rewards, such as research grants.
Merton’s reward system contains a set of rewards, a set of potential beneficiaries, and a set of awards linking
elements from these two sets, which can be conceptualized as an affiliation or 2-mode network. The Matthew
Effect can be formulated as a hypothesis on the dynamics of this network, predicting an increasing concentration
of awards among beneficiaries.
Casting a reward system in the mould of a 2-mode network, our attention is drawn to the similarities and
Abstract: differences among rewards. As in the example of the Nobel Prize, rewards vary with respect to their social
prestige. Organizations may try to increase the prestige of their rewards by selecting the same beneficiaries as
more prestigious rewards. However, rewards are sometimes targeted at different types of beneficiaries or reach a
particular segment due to the selection procedure. Segmentation among rewards will decrease the concentration
of awards, weakening the Matthew Effect.
This paper presents a visualization approach showing both concentration and segmentation. It positions rewards
and their beneficiaries in a plane showing segmentation, adding the value of awards as heights expressing
concentration. This yields a landscape, which evolves over time. The approach will be applied to the system of
rewards for visual artists in The Netherlands, 1984-2005.
Session: Visualization
2-mode , 3-dimensional visualisation , network dynamics , network landscapes , networks across time ,
visualization , reward systems
Accepted: Yes

103 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Rich DeJordy
Authors: dejordy@bc.edu
Boston College Organization Studies Chestnut Hill, MA 02467 USA
Title: Introduction to E-Net, a software package for the analysis of ego networks
This paper presents E-NET, a software package designed specifically for the analysis of ego networks. The
program facilitates the analysis, visualization, and comparison of ego-network data, including the calculation of a
number of standard ego network metrics. Specific capabilities include summarizing the demographic composition
of the network (e.g., % women alters, average income of alters, etc.), summarizing the heterogeneity of the
Abstract: network (Blau’s heterogeneity index, standard deviations, etc.), measuring levels of homophily (similarity of ego
to alters on attributes), measures of structural holes and brokerage, and generating relational crosstabs (e.g.,
how many ties occur between people of different religions, or how ties connect members of different political
parties to members of different religions). A comparison of E-NET with more traditional network analysis tools
(e.g., UCINET/PAJEK) for these analyses is also presented.
Session: egocentric network
Keywords: egocentric networks , social network analysis tools , methodology
Accepted: Yes

104 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

David Dekker
Authors: david.dekker@gmail.com
Erasmus University Rotterdam, Department of Econometrics, Rotterdam, The Netherlands
Title: Ordinate or Subordinate to Social Norms: Role Stress in different Simmelian Broker Roles
In complex organizations people experience many different and changing expectations. In organizational
behavior theory the concept of roles stress expresses the psychological consequence of facing multiple sets of
social expectations. Krackhardt (1999) shows that brokerage of Simmelian ties makes people more prone to
experience role stress. Simmelian ties are ties embedded in cliques, which develop their own set of expectations
The argument is that a broker of Simmelian ties inherently has to cope with different sets of expectations, which
lead to role stress. This implicitly assumes that a Simmelian broker holds a solitaire position and is always
subordinate to the different sets of expectations she bridges. In fact, Simmelian brokers need not be solitaire and
neither be subordinate to different sets of social norms. In case of two structural equivalent Simmelian brokers a
partner exists that faces the same sets of expectations. Hence, a possible ally exists that may help to coordinate
responses to different demands, as well as develop expectations in a mutual beneficial direction. In case of three
structural equivalent Simmelian brokers a clique of Simmelian brokers exists that all have incentive to develop a
dominant set of stable expectations among them on how to manage sets of expectations in the different cliques
of which they are all members. In this paper we empirically study the social influence structural equivalent
Simmelian brokers have on each other and the amount of role stress they experience. Data are collected in three
different complex organizations. Results suggest that a distinction between ordinate and subordinate Simmelian
brokers enhances development of theories on Simmelian ties.
Session: Intra-Organizational Networks
Keywords: attitudes , belief systems , brokerage , coordination , dependence , framing , trust , role stress
Accepted: Yes

105 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Mario Diani
Authors: mario.diani@unitn.it
University of Trento, Faculty of Sociology, Via Verdi 26, Trento 38100, Italy
Title: Weak ties vs. strong ties in civil society networks: interactions or social relations?
In the first part of this paper I review available concepts of civil society and suggest how looking at the weak vs.
strong ties distinction may illuminate some important aspects of the dynamics behind interorganizational
networks and citizens’ organizing. I discuss in particular two different versions of strong ties, one based on
frequency of interaction, the other, on the content of ties - in particular, the embeddedness of alliance ties in the
bods created by overlapping memberships. The second section introduces the British study, based in Glasgow
and Bristol, which provides empirical grounding to this approach. Finally, I show how different versions of strong
ties generate different structures of civil society, one highly centralized, the other, closer to a polycentric,
decentralized model.
Session: Civil Society Networks
Keywords: associations , civil society , inter-organizational , political networks
Accepted: Yes

106 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jana Diesner
Carnegie Mellon University School of Computer Science Institute for Software Research Center for
Computational Analysis of Social and Organizational Systems (CASOS) Pittsburgh, PA USA
Authors: Kathleen Carley
Carnegie Mellon University School of Computer Science Institute for Software Research Center for
Computational Analysis of Social and Organizational Systems (CASOS) Pittsburgh, PA USA
Harald Katzmair
FAS.research Network Analysis for Business & Enterprise San Francisco, CA USA
The morphology of a breakdown: How the semantics and mechanics of communication networks from an
organization in crises relate
Previous research suggests that the patterns of intra-organizational communication change during crises.
Additionally, network-analytic studies indicate that during organizational crises interpersonal communication
becomes intensified, diversified, and tends to by-pass formal chains or hierarchies of communication more
strongly. However, the connection between the semantics and the morphology of communication networks from
organizations in crises is not well understood yet.
In our project we investigate this possible relationship by studying e-mail networks. The data set we use is the
Enron email corpus. Our research is based on the assumption that communication networks are the place where
organizational culture and identity are created through discourse and the circulation of stories. We furthermore
Abstract: assume that the semantic and structural mechanism of this process change during crises. More precisely, for the
times of crises we hypothesize that
a) The network segmentation and cohesion of network clusters increase, because people engage in strategic
alliances and small groups with trusted others.
b) The interpersonal usage of antonyms increases, because antonyms are one way or indicator for establishing
and distinguishing identity.
c) The semantic entropy of communication networks decreases, because the discourse drifts towards polarized
ends of themes and issues.
In our presentation we report on our findings with respect to these hypotheses.
Session: word networks or semantic networks
Keywords: semantic networks , organizational crisis
Accepted: Yes

107 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jan Kornelis Dijkstra

University of Groningen / ICS Interuniversity Center for Social Science Theory and Methodology Grote
Rozenstraat 31 9712 TG Groningen The Netherlands Phone: +31-50-3636208 Fax: +31-50-3636226
Title: Being Liked or Disliked:A Dual Perspective on Acceptance, and Rejection of Popular Adolescents by their Peers
A central feature of adolescent’ peer groups is the relative ranking of members on the dimension of popularity.
Almost every peer group contains a limited number of adolescents with a popular status. Despite that popular
adolescents seem highly attractive for affiliation, they are not automatically liked (i.e., accepted), but also disliked
(i.e., rejected) by their peers (LaFontana & Cillessen, 2002).
However, findings from previous research are based on the aggregated scores for the whole sample.
Consequently, explanations for acceptance and rejection of popular adolescents by their peers only pertain to the
aggregated level. These studies do not explicitly test for the effects of characteristics and behaviors of popular
adolescents on their relations with peers on the dyadic level. Hence, it does not help us to explain why the same
popular adolescents are liked by some peers, and disliked by others.
The present study was undertaken to examine on the dyadic level under what conditions adolescents are more
likely to accept, or reject their popular peers. To this end, we address the effects of antisocial and positive
characteristics of both popular adolescents and peers by means of the p2-model, especially suited to analyze
relationships on the dyadic level (Van Duijn, Snijders, & Zijlstra, 2006). Data stem from about 50 networks in
school classes.

Word Count: 229 (without references)

LaFontana & Cillessen, LaFontana, K. A. & Cillessen, A. H. N. (2002). Children's perceptions of popular and
unpopular peers: A multimethod assessment. Developmental Psychology, 38, 635-647.
Van Duijn, M.A.J., Snijders, T.A.B., & Zijlstra, B.J.H. (2004). P2: a random effects model with covariates for
directed graphs. Statistica Neerlandica, 58, 234-254.
Session: Adolescent Friendship Networks
Keywords: adolescence , centrality , multi-level networks , social relationship , affiliation networks , peer networks
Accepted: Yes

108 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Marya Doerfel
Authors: mdoerfel@scils.rutgers.edu
Rutgers University, Dept. of Communication, SCILS, 4 Huntington St., New Brunswick, NJ 08901, USA
Title: Reconnecting and Rebuilding Organizations and their Social Networks Following Disaster
Hurricane Katrina caused wide-spread disaster on every level of civil society in New Orleans and its surrounding
areas. Physical destruction deleted whole neighborhoods, technological destruction cut phone contact and other
communications throughout the city, and social destruction broke apart the ties that bind. Although much is
understood about social networking and its ancillary benefits, in general, the specific case of New Orleans’
situation provides a unique opportunity to extend an understanding of the social processes associated with
building – and now, in New Orleans case, rebuilding – civil society. Unique to rebuilding the social capital is that
former trusted partners are not necessarily back, and for those who have returned, social, financial, and
informational resources vary greatly. This study extends social capital and social support theories to the context
of returning and rebuilding by organizational and business leaders following a natural disaster. The focus is on
the organization-level aspects of embeddedness and scope and diversity of networks in understanding stories of
return. Accounts about how leaders got their businesses and organizations back up and running while
simultaneously reconnecting and rebuilding their social networks show the advantages gained from a
combination of human, financial, and social capital. Since December 2005 (following the regional devastation
caused by Hurricane Katrina), field research and in depth interviews with over 50 organizational and business
leaders/owners have been conducted. Follow up interviews and new participants continue to be added with
ongoing field research trips to New Orleans.

*Funding for this study was provided by National Science Foundation grant # 0554959.
Session: Communication Networks
corporate social capital , disaster response , inter-organizational networks , interviews as data ,
organizations , qualitative network
Accepted: Yes

109 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Bethany Dohleman
University of Illinois at Urbana-Champaign, Quantitative Psychology Champaign, IL, USA
Harold D. Green, Jr.
Science of Networks in Communities (SONIC), National Center for Supercomputing Applications, University of
Authors: Illinois at Urbana-Champaign
Dmitri Williams
Department of Speech Communication, University of Illinois at Urbana-Champaign
Noshir Contractor
Department of Speech Communication & Science of Networks in Communities (SONIC), National Center for
Supercomputing Applications, University of Illinois at Urbana-Champaign
Title: Putting Some ‘WoW’ into Modeling Longitudinal Networks
This study models the co-evolution of communication networks and perceived expertise among members of a
Massively Multiplayer Online Role-Playing Game called World of Warcraft (WoW). Specifically, this study tested
multi-theoretical multilevel hypotheses about what motivates an individual to create expertise-seeking advice ties
from other players within their virtual communities, called guilds. Previous research found that, consistent with
Transactive Memory theory, guild members are marginally (although significantly) more likely to go for advice to
those who they perceive has expertise on a certain topic. However, based on their assessment of the advice they
received, members might recalibrate their perceptions of others’ expertise. Thus the members’ advice networks
co-evolve with the perceived expertise of other members.

In this study, building on our earlier findings, we focus specific attention on the dynamics of this co-evolution. We
used SIENA to test hypotheses about how co-evolutionary dynamics vary for WoW communities that have
multi-modal means of communication (audio and text) versus uni-modal means of communication (text only).
Further, we test hypotheses about how the co-evolutionary dynamics are themselves time-dependent – changing
more immediately after introducing new communication technologies than compared to later in the timeline.

Nine guilds received a VoIP (Voice over Internet Protocol) headset, allowing them to communicate with each
other using audio and text means. Changes in network structure and guild member attributes in these treatment
groups are compared with changes in seven guilds who served as controls, relying only on text means.

Results reveal that mechanisms associated with the co-evolution of network structure and perceived expertise
levels varied between treatment and control guilds. For example, experts in guilds who use VoIP are more likely
to identify and approach other experts for advice, supporting theories of homophily and suggesting that the
primary driver of guild formation is bonding. Other advanced modeling techniques reveal that, for guilds that use
VoIP, rates of communication are larger in the initial time interval than in other time intervals. Similar patterns are
not found in control guilds. This study compares the patterns of network evolution for control and treatment guilds
and demonstrates new statistical tools for investigating how communication strategies change over time.
Session: On-Line Communities
access to knowledge , actor-driven modeling , dynamics on networks , network evolution , siena , statistical
Accepted: Yes

110 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Silvia Dominguez
Authors: Department of Sociology Northeastern University 350 Huntington Ave Holmes 525 Boston, MA 02115
Isidro Maya Jariego
Departamento de Psicología Social Universidad de Sevilla Campus de Ramón y Cajal 41.005-SEVILLA (Spain)
Title: The Other Side of Acculturation: are immigrants integrated in the personal networks of host individuals?
Abstract. There has been a vast amount of research on the changes experienced by immigrants, but little is
known about the changes experienced by host individuals. This article focuses on the role of host individuals in
the networks of relations between immigrant populations and the communities from the dominant culture, as well
as the changes experienced by host individuals because of the continuous contact with immigrants. This
research applied a network approach to the study of the acculturation of host individuals. Two independent
studies were carried out: a systematic analysis of the personal networks of Argentinean (n = 67), Ecuadorian (n =
59), Italian (n = 37) and German (n = 37) residents in Seville and Cadiz (Spain) (Study 1); and an ethnographic
Abstract: study with human service workers for Latin American immigrant USA (Study 2). With two different strategies, the
role of host individuals in personal networks of foreigners in the United States and Spain was analyzed. The
results show that host individuals tend to have less centrality than compatriots, showing an overall secondary role
in the personal networks of immigrants. The lowest average centrality was observed in recent and temporal
migrants, whereas the highest corresponded to the individuals with more time of residence in Spain. The
personal networks of human service providers in USA vary in ethnic composition and in their structural
properties, and therefore shape different type of integrative bridges for immigrants.
Session Organized by Silvia Dominguez & Betina Hollstein
Session: Mixed Methods: Combining Qualitative and Quantitative Approaches
Keywords: acculturation , immigrants , personal networks , centrality , mixed methods , host individuals
Accepted: Yes

111 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Patrick Doreian
Authors: pitpat@pitt.edu
University of Pittsburgh, Sociology, Department of Sociology 2602 WWPH, Pittsburgh, PA 15260, USA
Title: Routes to and From the English Premier League
The movement of football (soccer) players between countries to ply their trade is an enduring phenomenon that
has lasted for over a century. These movements generate networks between clubs and countries that can be
studied in a variety of ways within different designs. The usual design considers teams representing their
countries in the World Cup Final and note where these players play their club football. Such data are two mode
network data and provide snapshots every four years of the links between countries. Such images obscure a
more complicated network of player moves between countries and clubs. Another framework is to look at
aggregate moves between the FIFA sanctioned regions to establish an overall image of the structure of
international football. Yet another complementary framework is to study all of the players for a single prominent
league over an extended period of time. Using the English Premier League (EPL) as such a football destination,
data have been assembled for the entire careers of players for two seasons (1992-3 and 2006-7) with a view to
examining the distribution of routes that players make to and from the EPL. These routes vary greatly in their
length (ranging from 0 to 28 moves at the latest count) taking the forms of paths, walks and treks between clubs
and through countries. These routes are located in a very complicated network for the clubs involved and a less
complex network for the nations through which these players move. Algorithms for identifying shortest and
longest paths are of little value in studying the movement of players in these networks. Some of the mechanisms
that generate these routes are discussed.
Session: International Networks
Keywords: social geography , trade , world systems
Accepted: Yes

112 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Joss Douthwaite
Health and Social Services, Bienvenue,, La Charriere du Bourg,, Grouville, Jersey JE3 9UN, Channel Islands,
Evaluating cross functional teams: comparing managerial expectations and team members’ perceptions through
use of SNA
How can government departments work together to solve complex problems? One approach is to set up a cross
functional team with members drawn from the various departments. However, these teams frequently experience
problems pooling their knowledge and skills and also in acting as brokers between the rest of the team and the
departments to which they belong.
The literature suggests these problems could be due to inflexible organisational structures (e.g. home
departments are too rigid) or competing professional and social identities. This study uses SNA to map out the
relational ties between professionals from several government departments and agencies who participated in a
Abstract: cross functional Youth Action Team on the island of Jersey, UK. The team deals with children identified as 'at risk
of offending' or who are already involved with the youth justice system. The purpose of the SNA was to identify a)
the commissioning managers’ expectations for the cross-functional team network (i.e. ‘ideal network’); b) team
members’ perceived ties between themselves 12 months after the team was set up (i.e. ‘actual network’); and c)
team members’ perceived ties between them and their home departments (i.e. team members' ‘ego networks’).
Results show that the relationships between team members and their home departments affects both the level of
integration within the team and the individual team members' ability to liaise between their home departments
and the rest of the team.
Session: Intra-Organizational Networks
Keywords: interdisciplinarity , intraorganizational networks , social capital , brokerage
Accepted: Yes

113 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Lydia Drumright
University of California, San Diego Department of Family and Preventive Medicine Division of International
Health & Cross-Cultural Medicine La Jolla, CA, USA
Susan Little
Authors: University of California, San Diego Department of Medicine La Jolla, CA, USA
Douglas Richman
University of California, San Diego Department of Pathology La Jolla, CA, USA Veterans Affairs Healthcare
System, San Diego, CA, USA
Simon Frost
University of California, San Diego Department of Pathology La Jolla, CA, USA
Using Social Network Methods to Analyze Phylogenetic Clustering of HIV among Recently-Infected MSM in San
Diego, California
Background: Molecular analysis of HIV transmission clusters can help elucidate risk factors for transmission at
the network level. Applying social network analysis approaches, we examined correlates of transmission clusters
among men who have sex with men (MSM) with recent HIV infection.
Methods: 285 MSM with recent HIV infection were recruited between June 1996 and 2006. Sequences of the
HIV-1 pol gene were compared to determine clustering, defined as less than 1% nucleotide divergence between
two individual’s viruses. Exponential random graph models were fitted to the cluster matrix to determine factors
associated with clustering between individuals.
Results: The mean sequence divergence between individuals was 5.6%; 90 individuals (32%) belonged to a
Abstract: transmission cluster. There was no significant change in clustering over time (odds ratio (OR)=1.02, p=0.15).
Individuals sampled further apart in time were less likely to cluster (OR=0.6/yr, p<0.01). Individuals were less
likely to cluster together if they were discordant by age (OR=0.96/yr, P<0.01) and more likely to cluster if they
had drug resistant virus (OR=3.0, p<0.01). There was a trend for individuals with acute infection to cluster
(OR=1.5, p=0.08).
Conclusions: Previous studies have considered correlates of belonging to a cluster, rather than belonging to the
same cluster. We overcame this potential ecological fallacy by treating a pair of individuals with similar HIV
strains as the outcome. Our results suggest that that there may be sub-epidemics with higher levels of drug
resistant virus; and that people with acute infection are more likely to transmit HIV to others.
Session: Infectious Diseases and Social Networks
hiv , clustering , epidemiology , ergm (p*) , sexual contacts , sexually transmitted infection , men who have
sex with men , hiv drug resistance
Accepted: Yes

114 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Andrea Edgar
Authors: andrea.edgar@usma.edu
USMA, West Point, NY
Title: Network Visualization for Analysis of Terrorist Organizations
Against conventional thinking, even the strictest organizations are not merely hierarchal systems. Many times
other connections exist that result in a more dynamic, complex, and intricate social network. Understanding how
people are interconnected through various channels can provide a great deal of information as to how a specific
network functions and operates. The advantage that a social network diagram has is that it can provide a
graphical representation of where key actors are located and what their role in the network is.
As Social Network Analysis is a growing science, little research has been conducted on how organizational
members interpret graphical representations of their own networks. Applying Chris Wickens’ Model (1988) and
Signal Detection Theory, this experiment focuses on the user's ability to correctly interpret and identify the "key
actors" in a social network diagram. Participants were evaluated on four different graphs which varied four
features of graphs: node size, edge weight, edge count, and distance from the center. Our research found that
inexperienced users identified key leaders based more on line thickness than node size. Therefore there is a
higher correlation between edge thickness and selection of key leaders on a graph than with any of the other
three aforementioned graph features.
Building on that previous research and experimentation, new graphic representations are being built and tested.
Applying Social Network Analysis to the analysis and identification of possible suppliers and detonators of
terrorist organizations is the context in which this experiment was conducted.
Using Gestalt, Tufte, and graphic design principles, five factors were identified as being possible contributors to
either inhibiting or facilitating graph readability: node size, node color, line color, line thickness, and contrast. By
varying one, or a combination of these factors, a participant’s ability to correctly interpret the data on the graph
varies. This research shows the effect that changing aspects of the graphic representation can have on
interpretation of the data.
Session: Visualization
Keywords: data representation , experiment , terrorist networks
Accepted: Yes

115 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Christofer Edling
Authors: christofer.edling@sociology.su.se
Stockholm University, Dept of Sociology, S-106 91 Stockholm, Stockholm, Sweden
Title: Local elites networks in Sweden
We study the patterns of formal and informal interpersonal affiliations connecting different categories of local
decision makers empirically, on the basis of theories grounded in the classics of elite studies, as well as in
contemporary network analysis. We sample five Swedish municipalities, and for each of these we collect detailed
ego-network data from approximately 30 persons and Complete” network data for a 100-node network.
The findings reported in this paper are the results of a pilot study conducted in a municipality located in the region
of Stockholm, in late spring 2006. The three main objectives of this limited study have been to (1) empirically
explore the methodological feasibility of network data-collection in a Swedish local elite setting, to (2) draw upon
these experiences in further fine-tuning our data-collection instruments and to (3) examine the observed network
structures for a preliminary discussion of substantial results. The study draws on interviews conducted with a
sample of seventeen local business leaders, politicians and the administrative leadership of the municipality at
hand. Two types of professional and private social networks have been studied: (a) the self reported, ego-based
networks of the interviewees and (b) the total networks of the interviewed elites. Results confirm the feasibility of
social network data-collection in Swedish local elite settings, and provide us with important insights regarding, for
instance, the optimal size of the sample population, and the design of the data-collection instruments. The
empirical results demonstrate systematic subdivisions within the studied elite, according to both their individual
characteristics and collective interests.
Session: Political Networks
Keywords: local elites
Accepted: Yes

116 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Kate Ehrlich
Authors: IBM, Research, 1 Rogers St, Cambridge, MA 02142, USA
Inga Carboni
Boston College Carroll School of Management Chestnut Hill, MA 02467
Building better teams by building better leadership: The role of instrumental and affective ties in the networks of
successful team leaders
Teams are a fundamental unit for organizing work in most enterprises. But as teams take on the appearance of
loose networks, reflecting the multiplicity of relationships and responsibilities within and without a single unit, the
role of the leader may need to be re-thought. Recent research on leadership has highlighted the importance of
building intra-organizational instrumental and affective relationships. Team leaders in particular are encouraged
to facilitate relationship-building among team members. However, little is known about how different types of
leader-member networks impact team performance. We studied leadership in 20 sales teams; 10 teams who had
high growth and 10 that were struggling by looking at their role across 8 networks half of which represented
affective ties and half of which represented instrumental ties. The leaders of the high performing teams differed
significantly from their counterparts in their role within the affective as well as their role in the instrumental
networks. There was no difference in personal performance.
Session: Leadership Networks
Keywords: business networks , leadership and networks , inter-organizational , coordination
Accepted: Yes

117 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Kenth Engø-Monsen
Telenor R&D, B6d, Snarøyveien 30, Fornebu 1331, Norway
Authors: Geoffrey Canright
Telenor R&D, B6d, Snarøyveien 30, Fornebu 1331, Norway
Valencia Remple
University of British Columbia, BC Centre for Disease Control Epidemiology, Vancouver, BC, Canada
Title: Topographic spreading analysis applied to a Vancouver sex workers’ network
We report on an application of the “topographic” analysis of spreading on networks to the empirically obtained
network of female sex workers (FSW) and their customers in Vancouver, Canada. The topographic analysis uses
eigenvector centrality (EVC) to define regions (subnetworks) in which spreading is fast and fairly predictable in its
progression, and has been well supported by simulations. This is however the first application to a sexual
network—furthermore, one for which the “starting” (currently) infected nodes are known. The data source is from
a network study conducted with indoor FSW in Vancouver, Canada. Interviews were conducted with 49 FSW
from seven indoor sex establishments. The interviews elicited proxy data on up to 10 of each of the respondents’
sex partners (first order partners), and proxy data on any known partners of their partners (for example, wives,
girlfriends, or other FSW). The resultant network was a fully connected component of 553 nodes with 2996 edges
(1% density). We will give the predictions, obtained from the analysis, for spreading of HIV—and (if time permits)
other STDs under varying assumptions. Also, we will use the analysis to propose specific preventive measures,
and their likelihood for retarding the spread of the infection.
Session: Sex, Drugs, and Social Networks
Keywords: canada , commercial sex workers , hiv , model-based simulation , sexually transmitted infection
Accepted: Yes

118 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Eliza Eranus
Authors: Corvinus University of Budapest, Sociology, Valyog utca 5 szam 3 emelet 11 ajtó, Budapest 1032, Hungary
László Letenyei
Corvinus University of Budapest Sociology
Title: Telehouse or private house?
The spread of personal computers and internet connections across households cannot be considered as a
typical diffusion of innovations process in Hungary contrary to western economies. The reason for this is that
these processes are highly influenced by governmental programs. One of the most important of these is
“telehouse” (later e-Hungary access point) program, which developed a network of public internet access points
across the contry, including the small villages.
This paper questions if “telehouse” program is the most effective way of increasing digital literacy. We compared
the “telehouse” program with placing personal computers and internet access to private households for certain
Abstract: period. To answer this question we made an experimental research in a Hungarian small village. We analyzed
the spread of computer and internet-knowledge from the point of view of relations and social network.
The results show that if people has personal computers and internet-connection placed at their home, the
diffusion of computer and internet-knowledge in their ego-network is more efficient compared to the case, when
they use the computers at the public access points. Our results show that that people tend to share their
computer and internet knowledge to their strong ties like their family and their best friends. This means that the
method of implementing computers and internet-connections in households gives an opportunity to decrease the
relatively high Hungarian indices of digital divide by the social connections between people.
Session: Public-Private network interaction
Keywords: experiment , personal networks , social capital
Accepted: Yes

119 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Bonnie Erickson
Authors: University of Toronto, Sociology, 725 Spadina Avenue, Toronto, Ontario M5S 2J4, Canada
Rochelle Cote
University of Toronto, Sociology, 725 Spadina Avenue, Toronto, Ontario M5S 2J4, Canada
Title: Varieties of Social Capital and Their Sources
One of the most important sources of social capital is the variety of people in a person’s entire social network,
including weak as well as strong ties. However, there are several kinds of variety, known to have different
consequences. This paper distinguishes five kinds of variety: (1) the number of different occupations in which the
focal person knows anyone, (2) the number of middle class occupations in which the person knows anyone, (3)
Abstract: the number of working class occupations in which the person knows anyone, (4), the number of occupations in
which the person knows a man, and (5) the number of occupations in which the person knows a woman. We
show both some common patterns and some variations in how these different kinds of social capital are related
to age, gender, education, work and family roles, voluntary association participation, and area of residence, using
the 2004 Canadian federal election survey.
Session: Social Capital
Keywords: canada , personal networks , social capital , inequality , gender
Accepted: Yes

120 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Emily Erikson
Authors: erikson@soc.umass.edu
University of Massachusetts, Sociology, Sociology Dept/SADRI, 240 Hicks Way, Amherst, MA 01003, USA
Title: Integration of Economic Systems: Interaction between Networks
Market expansion has been abstractly conceived as the expansion of one system, for example, the expansion of
capitalism from northern Europe across the world. In fact, market expansion is almost inevitably the integration of
two or more economic systems: the integration of the multiple economic systems into one European Union, the
opening of China to U.S. businesses, or the articulation between barter and cash economies on Lake Titicaca. I
consider the structural effects of different patterns of integration. Trade networks often have a core-periphery
structure. This implies six coarse categories of integration. For example, core actors may establish links with
Abstract: another system as in imperialist expansion; semi-peripheral actors may initiate trade; marginal figures may link
different trade networks, as was common in pre-modern commerce; or actors may link across types. I first
consider static models of integration by class of actor and the effect on network characteristics tied to the
distribution of goods and information within the network. I then move to a dynamic model of interaction, where
actors update their interaction preferences based on changes in the network structure created by previous actor’s
decisions. Finally, I build in additional complexity by considering how a constant flow of goods across the network
is affected by the emergent structures and is channeled into different distributions of resource allocation.
Session: Policy Networks and Governance
Keywords: trade , economic networks , dynamics on networks , globalization , world systems
Accepted: Yes

121 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Henrik Ernstson
Stockholm University, Departement of Systems Ecology, Stockholm, Sweden
Authors: Sverker Sörlin
Royal Institute of Technology, Division for History of Science and Technology, Stockholm, Sweden
Thomas Elmqvist
Stockholm University, Departement of Systems Ecology, Stockholm, Sweden
The role of social network structure in informal protection of green areas and urban ecosystem services in
Exploitation of green areas often reduces the capacity of urban ecosystems to produce local ecosystem services
(reduction of air pollution, noise reduction and recreational space). In Stockholm, a social movement has
succesfully protected a biodiversity rich central park of great size (27 km2) and forced management to operate
on a more fitting ecological scale. A social network analysis, based on a questionnaire to 60 organisations,
revealed a core-periphery structure where core members are cultural and nature protection groups with many
contacts to authorities, and where periphery members instead tend to be user organisations with many days in
the park (e.g. sports, boating and gardening groups). The structure integrates information from within and outside
the movement to core members, giving them a brokerage position which has enabled them to develop effective
methods for political action and to control the agenda of the movement. Periphery members secure legitimacy for
the political project as well as acting as sensors for what is happening on the ground. However, there are traces
of division; especially allotment gardens have formed their own component suggesting an ideological cleavage
between heavy use of the park and admiring it for its aesthetic values. The case points out structural network
factors to why social movements can shift power and transform governance systems and thus play an essential
role in bringing forward improved ways of managing urban green areas.
Session: Acivist Networks and Global Movements
Keywords: brokerage , governance , urban green areas , social movements , core-periphery structure
Accepted: Yes

122 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Martin Everett
Authors: University of East London London UK
Steve Borgatti
Boston College Organization Studies Chestnut Hill USA
Title: Induced Centrality
A graph invariant is a number derived from a graph that does not change if the nodes are re-labelled. A simple
example would be the number of nodes or the number of edges. In some circumstances we can use a graph
invariant to derive a centrality measure by determining to what extent each individual node contributes to the
overall graph invariant. We call this induced centrality. One method would be to delete the node and see how the
Abstract: graph invariant changes and use this as measure of centrality. Hence we first compute the invariant on the
original graph to obtain a value say K, we would recompute the invariant on the graph with a vertex x removed to
obtain the value P. The induced centrality of x is simply K-P. As an example take the graph invariant as the total
number of edges then the induced centrality turns out to be the familiar degree centrality. In this paper we
investigate the class of induced centrality measures.
Session: Mathematical Models
Keywords: centrality
Accepted: Yes

123 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Nikos Fakiolas
Authors: nfakio@ekke.gr
Greek National Center for Social Research Athens Greece
Title: Social networks in the Greek health-care system
Social capital consists of actual or potential resources that inhere within social networks or groups for personal
benefit (R. Carpiano, 2005). Several research efforts have been made to understand how community factors
impact health and well being. Some authors argue that social capital adds more to the physical and psychosocial
balance of the people than health manpower itself (R. Rose, 2000), since social networks reflect social
inequalities which influence health and the access to health services (J. Lomas, 1998). A considerable amount of
social capital is connected to lower premature mortality from cardiovascular diseases, but not from cancer (A.
Kimberly, , et.al., 2003). On the other hand, it is important to explore the characteristics of the people who seek
additional help from their social networks.
Part of a research project on health (sample aged 40 years and over) of the National Center for Social Research
addresses these questions. It provides several data, which connect social capital with the state of health of
specific social groups of the Greek population. Support to the people who need additional help is mainly offered
by individuals, relatives or professionals and rarely by organizations or other social institutions' officials.
Session: social networks and health
Keywords: health care providers , social capital , social capital and health
Accepted: Yes

124 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Tanja Falkowski
Title: Discovering Communities in Environments with Fluctuating Members
Communities are highly dynamic social networks whose structure changes over time. We observe communities
where members leave while new ones join and the community is still there even if all of its original members have
left. However, a community is typically defined at each point by its current members and the evolution is analyzed
by observing changes in their interaction behavior. Thus, if the membership of a community changes it is no
longer recognized as the same community.
To capture this phenomenon, we propose to define a community as a persistent structure in a graph of
interactions among fluctuating members. We find these communities by detecting the structural breaks between
them. Our method for community detection in changing environments involves three steps: At first we partition
Abstract: the time axis and cluster the interaction graph to find community instances of densely connected individuals. In
the second step we build a graph of similar community instances and apply a clustering to detect groups of
similar instances. In the third step we detect structural breaks that indicate periods when communities are
We evaluate our approach using data from an online community of international students. This community
exhibits, due to the fact that many students leave/join before or after semester breaks, a large amount of
fluctuating members and thus several structural breaks along the time axis. We show that we are able to detect
these significant changes that are caused by a high fluctuating membership and/or a highly changing interaction
Session: On-Line Communities
changes in networks , community structure , communication graphs , dynamics on networks , network
evolution , visualization , subgroup identification , longitudinal
Accepted: Yes

125 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Yiwei Fang
The School of Management, Xian Jiaotong University, No.28 Xianning West Road, Xian City,710049, Shaanxi
Authors: Province, China
Yang Zhang
Department of Industrial Engineering, The School of Management, Xian Jiaotong University, No.28 Xianning
West Road, Xian City,710049, Shaanxi Province, China
Title: Formation of Social Networks with Heterogeneous Actors
This paper studies a strategic network formation problem in the settings of heterogeneous actors. By cost-benefit
analysis, we reveal the features of efficient networks and stable networks, which arise when rational actors make
decisions to trade off the costs of forming links against the potential benefits from doing so. A distinct result
derived from analytical reasoning is that the actor with the lowest intrinsic utility plays a central role in static
equilibriums. Motivated by this finding, we propose an index called "utility attribute" to measure the degree to
Abstract: which the intrinsic value of the observed actor is lower or higher than that of the average level. Through a
simulation-based approach, we examine how utility attribute affects the efficiency and stability of network
formation. In the context of organizations, our findings in both analytical reasoning and simulation probe into two
issues below. (1) Why does a subordinate, rather than a superior, often play a significant role in informal
organizations? (2) To what degree a superior is rich in his capacity and resources can help the organization to
perform as its best?
Session: Mathematical Models
Keywords: network evolution , network structure , social network , economic networks , heterogeneity
Accepted: Yes

126 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dimitri Fazito
Federal University of Minas Gerais, Centre of Development and Regional Planning, Rua Alvaro Alvim 126Bairro
Authors: Alto Caicara, Belo Horizonte, MG CEP 30775, Brazil
Weber Soares
Federal University of Minas Gerais Department of Geography and Urban Planning
The qualitative network research applied to the international migration phenomenon: the case of irregular
Brazilian emigration to US
Social scientists have associated the causality chain of migration with the existence of social networks. However,
such association is only metaphorically apprehended. We discuss briefly some methodological possibilities of the
Social Network Analysis applied to the study of migration (Network of Flows Model; Full Migration Network
Model; Actor Network Model). Then we present a sensitive proposal application of qualitative network
methodology for the study of social networks in the migration process.
We applied Chris McCarty´s (2002) Personal Network model to the study of a typical case of international
emigration of Brazilians to US. In this qualitative research we interviewed 50 international returned migrants
residents in the Governador Valadares city (the Brazilian city with the largest proportion of families with
Abstract: international emigrants to US). Then, we reconstituted their personal networks looking for structural patterns (a
sort of typical topology) that might configure the international migration. We aimed the identification of the
complex brokerage system of mobster migration organizations resposible for the irregular flows to US and
particularity of such international migration system.
We present in detail the combination of qualitative techniques (daily ethnography and in-depth interviews) for the
production of "map networks" used in the displacement process. Such "symbolic maps" should represent the
structural pattern of displacements embedded not only in the physical but mainly in the social space of national
societies. At the end, we raise some issues on the validity of the application of qualitative network research for
the explanation of causality chain in the migration phenomenon.
Session: Qualitative Network Research: Methodological and Theoretical Issues
Keywords: intermediaries , migration networks , mental maps , personal network methods , qualitative network
Accepted: Yes

127 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Scott Feld
Purdue University, Department of Sociology and Anthropology, 700 West State Street, West Lafayette, IN 47907
Title: Rediscovering the Infrastructure of the Small World
In their original consideration of the Small World Phenomenon, Poole and Kochen speculated that severe
clustering could lead there to be great network distances between actors in different clusters that are far removed
from one another. Consequently, they were surprised by the results of their own thought experiments that
showed relatively short network paths even between farflung members of society. Milgram surprised many others
by empirically showing that people could actually find short paths through social networks that would connect
apparently farflung members of society. Other social scientists following them (e.g. Bernard and Killworth, Lin)
investigated the nature of those short paths, and what those paths suggested about the underlying network
structure that supported them.

Yet, their discoveries about the systematic nature of the short paths and the organization of the underlying
networks have been largely lost in recent rediscovery of the small world phenomenon. Watts proposed that
“random shortcuts” could make the connections underlying short paths, but there is no evidence that there really
are ties that approximate the properties of random shortcuts. Barabasi proposed that “sociometric superstars”
could make the connections underlying the short paths, but the Barabasi model also seems to assume
“randomness” among the connections of superstars to others.
This paper revisits sociological discoveries about the systematic (non-random) organization of large scale
societal networks that underly short societal paths traversing those networks. Specifically, we suggest that
societal networks can often be described in terms of a core and periphery (c.f. Borgatti and Everett) such that a
short path between any starter individual and any target individual can be constructed from three parts, 1)a short
path from the starter to some member of the core, 2)a short path from that core person to a second core person
closer to the target, and 3)a short path from that second core person to the target individual. Thus, we suggest
that it is often useful to decompose a large scale societal network into a core and periphery such that: a) Every
member of the core is connected to every other member of the core by a short network path; and b) Every
member of the periphery is connected to some member of the core by a short network path. It follows that every
member of society is connected to every other member of society by no more than two short paths from the
periphery to the core and one short path across the core.

We consider theoretical reasons to expect such a societal core-periphery structure, methods for finding the core
and periphery within a network, and further implications of
this type of core-periphery structure for such things as predictability of diffusion, network stability over time, and
robustness of measurability over missing data.
Session: Small World Research
Keywords: social geography , network structure , core-periphery , small world , horizontal/vertical networks
Accepted: Yes

128 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Thomas Fent
Authors: thomas.fent@oeaw.ac.at
Austrian Academy of Sciences, Vienna Institute of Demography, Vienna, Austria
Title: On the Emergence, Stability, and Replacement of Social Norms in a Dynamic Network
How individual behaviour is determined or at least influenced by social norms is one of the classic questions of
social theory. We consider a norm as a rule guiding individual decisions concerning rituals, beliefs, traditions, and
routines. Whenever coordinated behaviour is enforced without the help of an authority, this may be due to social
norms. The individual being in the situation of taking a decision at the micro level is guided by social norms
imposed at the macro level. The set of all individual decisions in a society generates the macro level behaviour of
the system which may strengthen or weaken the existing social norms. Thus, the long run development of social
norms is the result of collective dynamics within a social network.

We use an agent based simulation model to investigate the emergence, stability, and replacement of social
norms within a population of artificial agents. A social network connecting the agents serves to communicate the
social norms and the actual behaviour among the agent population. The agents in the network possess two types
of links connecting them with their ingroup and with their outgroup, respectively. Agents have the desire to be
associated and accepted by the members of their ingroup and they want to be different from the members of
their outgroup. Consequently, they derive a utility from adhering to the social norm of their ingroup and from
deviating from the social norm of their outgroup. Agents may adopt their behaviour according to the norms given
by their ingroup and outgroup and they may replace members of their ingroup and outgroup. Thus, our model
explains under what conditions social norms prevail within a subgroup of the society or even become global
norms being respected within the whole population.
Session: Collective Actions and Social Movements
Keywords: network dynamics , social networks , social norms , agent based modelling
Accepted: Yes

129 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Paulo Figueiredo
Romana Xerez
Miguel Mira da Silva
The Impact of Computer Based Communication on the Emerging of a Social Structure: the Analyses of two Case
Studies in Organizations
How does computer based communication contribute to networking in
organizations? We propose to analyze the influence of this network on
problem solving in organizations. A Pre-Internet analyses based on physical
contact, often around the coffee machine, is becoming less and less
important. Distance do not matter and many organizations are now spreading
globally, using outsourcing and offshoring, and taking advantage of Internet
technologies to "virtualize" the office. Many co-workers meet every day on
Skype but never saw each other. Many others work from home. Some companies
don't even have a physical office. Our data compares two different case
studies in organizations. We used software to extract data automatically
from the mail server, telephone exchange and messaging server. This data was
stored on a relational database according to SNA. Reports were produced
using the standard database reporting and analysis tools. We propose to
analyse how workers exchange information with co-workers, friends, relatives
and significant other using emails, telephone calls and messaging chats.
This paper describes the emerging of a Social Structure and we expect this
analysis will detect problems and come up with solutions to help the
organization exchange information quicker and better.
Session: Intra-Organizational Networks
Keywords: social networks , structure of collaboration , intraorganizational networks
Accepted: Yes

130 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Karen Fisher
Authors: The Information School of the University of Washington Box 352840, Seattle, WA USA 98195-2840
Tammara Turner
Community Technologies Group Microsoft Research Redmond, WA USA
Title: Qualitative Network Research: Methodological and Theoretical issues.
Women in the United States spend 4-7 trillion dollars annually and make 80% of household consumer decisions.
Scant research, however, has examined women’s role as family shopper. This study explored the importance of
women’s social networks in their role as shopper for non-grocery items. Using Fisher’s information grounds
theory (informal information flow in social settings), interviews were conducted with fifteen women ages 18
Abstract: through 70 for two hours followed by one hour of participant observation as they shopped in a large urban mall.
This talk addresses: (1) issues with over-reliance on a single protocol for identifying network data, (2) the need
for including informal and formal social settings as an alter focal point, and (3) why traditional hard social network
metrics are un-needed when using a qualitative perspective with deeper interest in understanding the intrinsic
nature and meaning of relationships, and (4) ways of bolstering "trustworthiness" (in Lincoln and Guba's terms).
Session: Qualitative Network Studies I
Keywords: methodology , qualitative network , qualitative research
Accepted: Yes

131 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Andreas Flache
Authors: University of Groningen, Department of Sociology, Groningen, The Netherlands
Michael Mäs
University of Groningen, Department of Sociology, Groningen, The Netherlands
The later, the better? Modeling effects of relationship timing on social cohesion in demographically diverse
Lau and Murnighan’s faultline theory argues that strong demographic faultlines in work teams may undermine
team cohesion. The theory implies that if demographic differences between group members are correlated
across various dimensions, the group is likely to fall apart into segregated subnetworks separated by strong
opinion differences. Our paper proposes a rigorous formal and computational reconstruction of the theory.

We integrate four elementary mechanisms of social interaction, homophily, heterophobia, social influence and
rejection into a computational representation of the dynamics of both actors’ opinions and their social relations.
Computational experiments demonstrate that that stronger demographic faultlines elicit more network
segregation aligned with more opinion polarization, consistently with faultline theory.
We then study effects of “relationship timing” (Moody 2002). While Moody has proposed that relationship timing
may greatly affect diffusion dynamics, we show that it may also help to temper the negative effects of strong
demographic faultlines in work teams. Computational analyses reveal that negative effects of strong faultlines
critically depend on who is when brought in contact with whom in the process of social interactions in the group.
More specifically, we demonstrate that faultlines have hardly negative effects when groups are initially split into
demographically homogeneous subgroups that are merged only when a local consensus has developed in each
subgroup. Contrary to other prominent theories of intergroup relations (e.g. contact theory) this suggests that
more intergroup contact may deteriorate intergroup relations under certain conditions.
Session: Intra-Organizational Networks
actor-driven modeling , model-based simulation , network dynamics , simulation study , relationship timing ,
demographic faultline , teams
Accepted: Yes

132 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Nikos Fokas
Authors: fokasz@ludens.elte.hu
Eotvos Lorand University, Department of Social Relations, Budapest, Hungary
Title: Political clientelisme as scale free network, Electoral behaviour in Epire
In my paper I would like present an application of scale free networks in the field of political sociology. From the
researches of Statis Damianakos we know that in Epire the influence of deputies who have a widespread politica
clientelist network forms a particular shape. As he noticed some candidates was able to obtain more than 80 per
cent of votes in their natives’ villages and from there their influence decrease gradually and concentrically.

However a strict distance dependence - supposed by Damianacos - can't be verified. In contrast I observed that
there is a relation between a number of villages and a level of influence, and we can easily conclude that in some
case the electoral results present a special structure. This result is very important because the above mentioned
dependence is the same which were observed in the cases of scale free networks between the scaling factor and
the number of pieces.

Since the interpersonal relationship of clientelism has obviously hierarchical character we can easily support that
the votes of Averof as well as these of Fontzos have a self-similar character. It is very important that this shape
of votes connect only to politicians who have a widespread political clientelist network. An other candidate
Derdemezis for example who without this traditional clientelisme had in this period also a great political influence
didn't present this structure
Session: Small World Research
Keywords: political networks , scale-free network , small world
Accepted: Yes

133 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Janna Fortuin
Authors: jfortuin@fsw.leidenuniv.nl
Leiden University Department of educational and child studies Leiden, the Netherlands
Title: Selection and influence of high school social network members: do goals matter?
Whenever the development of social networks of high school students is studied, it is apparent that dynamics of
behaviour are closely related to the dynamics of the networks. Certain characteristics of students might make
them more attractive to others (selection effects), but concurrently, the students might chance certain
characteristics because of their peers (influence effects). Both selection and influence effects have been
demonstrated in high school samples with regard to for instance alcohol consumption or music taste. Within
educational studies, there has been an emphasis on more educational characteristics such as goal directed
behaviour, and coordination of students’ goals and goals enforced (or encouraged) by the school. It is unclear
what, if any, role peer networks play in this goal endorsement and what role goals play in the development of
these networks. In this project, we combine both educational theories and social network studies to study both
goals and network development. A first analysis is presented of six (8th grade) classes in one Dutch high school.
Measurements of both goal endorsement and networks (in the form of students liking each other) within
classrooms were taken during three waves across one school year. Using SIENA, we model both the influence
the network has on goal endorsement of individual students and the selection processes that occur during this
school year.
Session: Adolescent Friendship Networks
adolescence , dynamics on networks , friendship networks , network coevolution , peer networks , selection
effects , siena , educational goals
Accepted: Yes

134 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Pacey Foster
Authors: pfoster@aptima.com
Aptima, Inc, 12 Gill Street, Suite 1400, Woburn, MA 01801, USA
Title: Beat research: Citation networks in sample-based music production
In rap music, producers routinely sample segments of other songs to construct new compositions. As producers
select from musical back catalogs for raw materials, they generate networks of citations between new and older
works. While reference and citation is common in many artistic forms, the widespread commercial reuse of
copywritten material was an early indication of the transformative impact that digital technology has had on
creative industries ever since. Long before music downloading, sample-based music production pushed the
boundaries of legality and generated much discussion about the political, legal and artistic implications of digital
music production. Despite the fact that these practices generated extensive network data on the relationships
between sampled songs and the songs that reference them, formal network research methods have not been
applied to understanding sample-based music production. Using a collaboratively generated data base of rap
music samples, I present the results of an exploratory study that shows how sample-based production generates
networks of citations with surprising non-random characteristics. In particular, the status rankings provided by
citations to original works are not well explained by traditional variables like mass market popularity. By exploring
the networks of citations among songs via samples, we can gain a deeper understanding of the social practices
underlying this important art form.
Session: Networks in Cultural Industries
Keywords: cultural industries , community of practice , citation network
Accepted: Yes

135 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Stephanie Fox
Authors: Simon Fraser Univerdsity, School of Communcations, Vancouver, Canada
Julia Wilson
Simon Fraser Univerdsity, School of Communcations, Vancouver, Canada
Title: Why Aren’t They Stealing My Ride? A Social Network Analysis Examination of Bait Cars
Auto theft in Vancouver, Canada has been on a steady decline since 2004, the year the province’s police force
and its partners inaugurated the bait car program, whose success includes a 25% reduction in auto theft since its
inception. In the program, bait cars, equipped with surveillance technology, GPS positioning transmitters, and
remote controlled disablers, are placed strategically in areas known for auto theft, effectively serving as bait for
would-be car thieves. Another major component of the program is a province-wide awareness-raising campaign
that runs across multiple media platforms. The RCMP, Canada’s federal law enforcement agency, credits the bai
car program with deterring “entry level youth” from stealing vehicles, while those caught by the program are the
most wanted car thieves of organized car theft networks in the Lower Mainland of British Columbia. The RCMP
believes these networks are behind a massive increase in the number of exported stolen cars over the last
decade, and that they are linked to other organized crime networks across Canada and internationally,
particularly in the developing world where limited supply of vehicles drives demand for organized auto theft.
This study examines the structure of these networks in the Vancouver area and the various relations between the
individuals involved. It relies on data gathered from police records and from interviews with police and
government officials as well as businesses associated with the automobile industry. Its shows the effect of the
bait car program on the structure and nature of these organized crime networks.
Key words: Bait cars, organized crime networks, auto theft
Session: Criminals, Gangs, Terrorists, and Networks
Keywords: organized crime networks , criminal networks
Accepted: Yes

136 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Markus Franke
Institute of Information Systems and Management Universität Karlsruhe (TH) 76128 Karlsruhe, Germany
Authors: Bettina Hoser
Institute of Information Systems and Management Universität Karlsruhe (TH) 76128 Karlsruhe, Germany
Jan Schröder
Institute of Information Systems and Management Universität Karlsruhe (TH) 76128 Karlsruhe, Germany
Title: Enlarging personal networks through transitive clustering
A phenomenon in social networks is the transitive closure, i.e. the question whether for
instance a friend of a friend can also be (or become more easily) a friend of mine. We study the
prediction of such closures - or parts thereof - without the need for parameter estimation. Our
results are based on an analysis of the EIES data set that contains the state of a researcher
network at two distinct points in time. For the analysis, we use the restricted random walk
cluster algorithm to detect related pairs of persons that share no relationship at the beginning.
Abstract: The algorithm scales well for large data sets and generates a hierarchy of overlapping clusters;
this allows asymmetrical inclusions of persons in each other's clusters, depending on the
direction of the original relation. The relations found in the clusters are evaluated against the
newly established relations contained in the data set's second part, where the share of relations
found correctly by the algorithm is used as precision measure. A good precision of about 86% is
achievable, but only for a small part of the data set. When using larger clusters, more relations
are found - at the expense of the results' quality.
Session: Methods and Statistics
clustering , network change , network dynamics , network evolution , networks across time , personal
Accepted: Yes

137 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Andreas Frei
Authors: Institut for Transport Planning and Systems (IVT), HIL F 36.2, ETH Hönggerberg, CH-8093 Zurich, Switzerland.
Timo Ohnmacht
Department of Sociology, University of Basle, Petersgraben 27, CH-4051 Basel, Switzerland.
Title: Social Geography and Job Mobility: Whose Private Life causes more Work?
Economies of scale, economies of scope and infrastructures like highways or airports provide us the possibility
for an economical growth, which leads to bigger markets. But bigger markets cause also bigger professional
activity spaces, which changes the requirements and needs of the society toward an ego and vice versa. These
requirements are highly linked to social networks, especially to their geographical dispersion.
This paper addresses the size and the structure of egocentric social networks focusing on how the geography of
social networks depends on events in the mobility biography and on sociodemographic factors. We investigate
the influence of the events during the life course, such as change of job, relocation and school career on the size
Abstract: of egocentric network. The experimental goal was to visualize the egocentric network geography and calculated
a measure for each case in the dataset. Within the given activity space, as a result of the individuals mobility
biography, the interest is also directed toward the organisation of social networks in terms of physical travel and
meetings through ICT technologies. The network geography figure for each dataset represents the “social activity
space” of an actor and is the dependent variable in multivariate data analysis. Therefore, the question to be
asked is whether the fulfilling of mobility chances and requirements during the course of life (e.g. job seeking)
induces geographically far-flung networks. The evidence shows that young and highly educated male actors with
frequent relocations are confronted with dispersed network geography.
Session: Egocentric networks and job seeking
Keywords: job seeking , inter-individual , social capital , social geography , egocentric networks
Accepted: Yes

138 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Thomas N. Friemel
University of Zurich, Institute of Mass Communication and Media Research, Andreasstrasse 15, Zuerich, ZH
8050, Switzerland
Title: The network structure of the knowledge gap
The knowledge gap hypothesis is an intensively discussed topic in mass communication science. It assumes that
the information gap between well and poorly informed people widens when additional information is provided.
This holds true for political information as well as other topics which are important for individuals and the society
as a whole. Traditionally the empirical testing of the hypothesis focuses on factual knowledge. This study uses
another approach by analysing cognitive associative networks. Data were collected using a representative panel
survey with three waves (N=1’068). Subject of the survey was a national referendum about a new foreigner and
refugee law in Switzerland which took place in September 2006. Respondents were asked for three spontaneous
associations on the subject of foreigners. These three items are treated as complete triads. To analyze “the
network structures of the knowledge gap” these individual cognitive networks were aggregated to bigger
networks by combining them according to various criteria’s (SES, pros and cons etc.). This presentation
discusses which structural characteristics these cognitive associative networks have, how they differ with regard
to different social groups and how they changed during the referendum campaign.
Session: Cognitive associative networks
access to knowledge , associations , belief systems , changes in networks , cognitive networks , immigrants ,
mental maps
Accepted: Yes

139 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Simon Frost
University of California, San Diego, Pathology, UCSD Antiviral Research Center, 150 W Washington St, San
Diego, CA 92103, USA
Title: Modeling respondent-driven sampling of social networks using stochastic context-free grammars
Respondent-driven sampling (RDS), a kind of chain-referral sampling, is becoming an increasingly popular
approach of sampling 'hidden' populations, such as injection drug users and men who have sex with men. RDS
involves giving study participants a small number of coupons to give to other potential participants who are their
friends or acquaintances, and employs a 'dual-incentive' system, where individuals receive compensation for
recruiting others, in order to sustain the recruitment process. As a side-effect, RDS provides information on
mixing between different populations and, by asking individuals about their relationship to the person that
recruited them, the extent of overlap between social and sexual networks. Current analytical techniques treat the
Abstract: recruitment process as a Markov chain, which is inappropriate as individuals may recruit more than one
individual. We show how stochastic context-free grammars (SCFGs) can be used to model the tree-like
recruitment process, which allows us to test for non-random mixing between subpopulations (e.g.
infected/uninfected), for independence of characteristics between recruitees of a given recruiter, and for
differences in patterns of mixing between different populations in a likelihood-based framework. We also
demonstrate how simulations of SCFGs can be used to design strategies to target specific subpopulations,
through offering increased compensation for recruiting these subpopulations and/or more coupons to individuals
in the subpopulation.
Session: Sampling methods
Keywords: social network analysis tools , respondent driven sampling
Accepted: Yes

140 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jan A. Fuhse
Authors: jan.fuhse@sowi.uni-stuttgart.de
University of Stuttgart Institute for Social Sciences, Unit IV Stuttgart Germany
Title: The Culture of Social Networks
The presentation discusses networks as socio-cultural structures, with special emphasis on the cultural side. As
conceived by Weber, Schütz, and Luhmann, all social structures are inherently cultural in that they are based on
meaning. Following authors from von Wiese and Elias to Fine and White, networks are configurations of social
relationships interwoven with symbolic material.
Based on these various theoretical strands, the following propositions are advanced:
1. Social relationships are dynamic structures of reciprocal (but not necessarily symmetric) expectations between
alter and ego. Through their transactions, alter and ego construct an idiosyncratic ‘relationship culture’
comprising symbols, narratives, and relational identities.
2. Culture is intersubjective and thus rooted in social networks (rather than merely subjective).
Abstract: 3. Categories are inscribed in the culture of social networks.
4. Actual transactions are the result of this cultural pattering of networks.
5. The distinction between cultural and structural properties of a network is analytical rather than substantial.
6. Both quantitative and qualitative methods are necessary to analyze the the interplay between the cultural and
the structural levels of social networks. Qualitative network research is able to deal more thoroughly with the leve
of meaning.
7. Important topics for research include: Investigating the micro-processes of relationship formation, of network
strategies and dynamics, the coupling of social relationships in a network, the cultural imprint on network
structures (for example through role categories or cultural blueprints for relationships and network structure), and
the intersubjective construction of identities in networks.
Session: Qualitative Network Research: Methodological and Theoretical Issues
Keywords: identity , methodology , social network , network culture , social relationship
Accepted: Yes

141 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Yoshi Fujiwara
NiCT / ATR, CIS Applied Network Science Lab, Hikaridai 2-2-2, Seika-chou, Souraku-gun, Kyoto 619-0288,
Title: Large-scale structure of production network and chain of firms bankruptcies in Japan
Production network refers to a line of economic activities in which firms buy intermediate goods from upstream
firms, put added-values, and sell productions to downstream firms, or consumers in the end of the line. The
whole set of these processes of putting added-values in turn forms a giant network of production ranging from
upstream to downstream, down to consumers. Net sum of added values in production network is basically the ne
total production, that is, GDP. Although production network, its structure and temporal change, are crucial in
understanding macro-economic dynamics, little has been hitherto studied by real data. Here we study a
Abstract: nation-wide production network comprising one million firms and links among them in Japan. Also we point out
that each link is usually a creditor-debtor relationship. If a firm goes into financial insolvency state or bankruptcy,
then firms on its upstream can have secondary effect from the bankruptcy. By using the data of bankruptcies, we
show that such network effect for bankruptcy is by no means negligible. In fact, nearly 20% of total debt is due to
such effect. Moreover, the link effect dominates, in probability, other causes for bankruptcy, such as poor
performance in business, for larger debt when bankrupted. We argue that the effect due to bankruptcy chain is
considerable in its ripple effect due to the heavy-tailed degree distribution.
Session: Networks, Economics, and Markets
Keywords: business networks , contagion , economic networks , productivity , scale-free network , trade
Accepted: Yes

142 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Crystal Fulton
Authors: crystal.fulton@ucd.ie
UCD School of Information & Library Studies, University College Dublin, Belfield, Dublin 4, Ireland
Title: Leisure, Learning, and Linking: The Social World of Irish Lacemakers
This paper explores the social world of older adults involved in the hobby of lacemaking and the connections they
develop within and beyond this community. In Ireland, the government has recently identified older adults as an
information poor group at risk for social exclusion (Information Society Policy Unit, Department of the Taoiseach,
2002). The Library Council of Ireland's Joining Forces report (2000), a landmark study of Irish culture and the
future of Irish libraries in meeting the needs of the public, and Ireland’s National Economic Social Forum
(www.nesf.ie), have similarly prioritized development of social inclusion for older adults.

Irish lacemaking, once a common and important means of economic support in Ireland, is now promoted and
practised in lace guilds as a leisure activity. Lacemaking may be characterized as serious leisure, that is an
amateur or voluntary activity in which the hobby forms a central life interest, with participants actively acquiring
Abstract: and expressing special skills, knowledge and experience (Stebbins, 1996; 1997). Information skills development
and information exchange are key features of lacemaking, with a variety of information and technical skills
needed to navigate the complex maze of resources which support this hobby. Lacemakers are frequently older
women, who are often retired and have time to devote themselves to hobbies.

A group of Irish lacemakers were observed as they interacted at lacemaking events hosted by their lacemaking
guild. From this group, twenty participated in interviews about their lacemaking, social connections within and
beyond their guild, and the role of lacemaking in their lives. By examining social connections in the lacemaking
community, we can learn about the various links lacemakers form with one another and how these links support
their information seeking behaviour. This analysis may help us understand the processes of learning and civic
involvement and inclusion through the hobby of lacemaking.
Session: Networks of Public Engagement and New Forms of Participation
leisure , inter-individual , social capital , information seeking behaviour , social worlds , older adults , social
Accepted: Yes

143 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Simone Gabbriellini
Authors: simone.gabbriellini@sp.unipi.it
University of Pisa, Department of Social Sciences, Pisa, Italy
Title: Networks and Games: a MMORPG guild analysis.
The basic hypothesis of the research is that virtual communities produces “real” relational ties, and game
communities are a privilegiate place to observe this process: in fact, people often access a game’s community
just because they like to play to that game; after, they play the game and they access a guild, meeting people.
The anecdotal evidence is clear: people prefer to switch game and continue to play with “friends” rather than play
the old loved game with unknown people.
So I say that MMORPG communities, or better guild communities, migrate from a game to another. I want to
investigate the way this happens. Using Social Network Analysis to analyze the information flows, but also using
agent based computer simulation to reproduce (and understand) the migration mechanisms.
This time I will use the online forum, in which they discuss their own opinons about the game and their personal
life, to understand how the structure of that network is made up and who are the people that play a central or
strategical role in the community.
I have longitudinal data (more or less 3 years), so I can see the evolution of the relationships in the community in
that period.
My aim is to reconstruct the relationship between people using the messages they had exchanged in the forum; I
will then analyze the dinamic of the network using Social Network Analysis to understand the evolution of its
Session: virtual communities network
Keywords: internet/hyperlink networks , mmorpg network
Accepted: Yes

144 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Cristina Gagliardi
I.N.R.C.A.Italian National Institute of Research on Aging, Research Department, Ancona, Italy
Anna Vespa
Authors: I.N.R.C.A.Italian National Institute of Research on Aging, Department of Neurology, Ancona, Italy
Roberta Papa
I.N.R.C.A.Italian National Institute of Research on Aging, Research Department, Ancona, Italy
Simonetta Rossini
Oncologist, Associated Hospital Umberto I-GM Lancisi-G Salesi, Ancona, Italy
Title: Social support networks of women suffering from early stage breast cancer
The aim of this study was to investigate the areas of depression, anxiety and social support using the structural
model of the social network. By comparing the networks of two samples of women, healthy and with early stage
breast cancer, it is possible to identify differences in their relationships, in the shape of the networks themselves
and the correlation with the levels of depression and anxiety. The results showed that the levels of anxiety and
depression were higher in women with breast cancer and that the variables of social network and social support
differently correlate with depression and anxiety in the two groups.
Session: social networks and health
Keywords: social network , social support , breast cancer , anxiety , depression
Accepted: Yes

145 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Cristobal Garcia
Authors: crisgh@mit.edu
MIT Comparative Media Studies,Cambridge,MA,US Columbia's Network Architecture Lab, NYC, US
Title: Mapping and Visualizing Organizational Electronic Networks in a Large European Accounting Firm
This paper consists on the analysis and visualization of organizational email networks of one hundred and fifty
(150) partners from all practices (A, B, C and D) in a large accounting firm of around 2,000 people distributed in
six (6) cities across a small European territory over a period of three months, i.e., from February to April of 2005.
We introduce and use TeCFlow, a Java-based software that analyzes and visualizes social network data from
email archives by building movies of them, illustrating their change across time, and by providing communication
frequency, number of emails among nodes, network position and role as well as measures of density and
centrality (degree and betweeness).
We found evidence that partners from practice A communicate more across linguistic boundaries than members
of practice B, C, and D do, that the type of practice/department matters more than the location (city) in terms of
information flows, and that C partners are more dispersed across the network. C partners exchange more emails
with B partners than with A partners. Overall, this organization’s network has both low betweeness and degree
centrality, and we observe a decreasing density towards the end of each month analyzed. We provide tentative
hypothesis for this behavior and network change. We also found that partners receive more emails than they
send out. However, there are clearly different individual patterns of email communication, and thus partners
(nodes) that might be considered “influentials, brokers, or gatekeepers” according to their centrality/peripherality
and position within the network. Finally, we summarize the findings and we discuss the overall characteristics of
this organization’s network ecology and its recurrent patterns as well as some privacy implications, and
guidelines for further work.
Session: Intra-Organizational Networks
changes in networks , business networks , central actor characteistics , collaboration , core-periphery ,
Accepted: Yes

146 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Livia Garcia
Authors: lgarcia@uma.es
University of Malaga (Spain)
Title: Networks and Identifications: the diffusion of a European identity through personal networks
In the study we have carried out the assumption taken is that the identification with the European Union can be
considered an innovation. Unfortunately, the network data that have been collected (Málaga, Spain, 2004)
pertain to only one moment in time, and consequently it is not possible to carry out this analysis. It is possible,
nevertheless, to study how at a particular moment in time European identity is distributed in a certain social
The general objectives of this research can be reduced to two:

1) To study the importance of the individual’s personal network in the development of an attitude towards the EU
and in the configuration of an identification with the Union and a feeling of European identity.
2) To study, within the personal network, the process of interpersonal influence, detecting whether there is a
leadership of opinion on Union issues and analysing the characteristics of the leader.

The general hypotheses of the study are:

(i) A person’s attitude towards the European Union and identification with it tends to be strengthened or
weakened within their personal network.
(ii) The people who are most familiar with the Union and have more information about it exert a greater influence
than the rest of the members of the network to which a particular subject belongs.
(iii) The people who first identify with the Union occupy central positions in the social structure, and this
identification is gradually diffused towards more peripheral positions.
Session: Networks and Identifications
Keywords: identity , diffusion , core-periphery
Accepted: Yes

147 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Kaberi Gayen
Napier University, School of Accounting, Economics and Statistics, 219 Colinton Road, Edinburgh, Mid Lothian
EH14 1DJ, UK
Tarun Kanti Gayen
Executive Director, Society for Communitu-health Rehabilitation Education and Awareness (CREA),Dhaka-1207,
Robert Raeside
Napier University, School of Accounting, Economics and Statistics, 219 Colinton Road, Edinburgh, Mid Lothian
EH14 1DJ, UK
Title: Understanding Network Norms of Heroin Users in Dhaka City, Bangladesh
What binds and endures small heroin user groups have been explored in this paper. Data were collected mainly
using a structured questionnaire and in-depth interviews of around 20 small groups [n=150] in central Dhaka,
where the HIV prevalence rate among the drug users, especially injecting drug user [IDU], is highest in
Bangladesh. Analysing social network data along with socio-economic and demographic data it was found that
an enduring network norm which is supported by physical settings of drug using that keeps ‘drug use ritual’ of
heroin smoking uninterrupted, is the main reason of using heroin in groups. Besides, central actors’
characteristics, mutual benefits like cost sharing, getting the access to substance even at the withdrawal phase
due to lack of money, helping each other when law enforcement agencies intervene, geographical proximity of
living arrangements of the users, and socioeconomic similarity were found to be important binding factors for
these small heroin user groups.
Session: Sex, Drugs, and Social Networks
Keywords: drug use , central actor characteistics , social networks
Accepted: Yes

148 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Patrick Gerland
United Nations, DESA/Population Division, New York, USA
Susan Watkins
University of California-Los Angeles, California Center for Population Research, Los Angeles and University of
Pennsylvania, Population Studies Center, Philadelphia - USA
Daily Conversation Logs and Diaries in Rural Malawi: New Insights about Topics and Patterns of Social
This paper uses non-traditional data collected in rural Malawi to examine the characteristics of social networks in
which villagers talk with friends, relatives and neighbors about AIDS.
Demographers have long known that such informal conversations provide information on topics such as family
planning methods and on diseases, including AIDS. Analyses of longitudinal survey data collected in rural sites
by the Malawi Diffusion and Ideational Change Project (MDICP) has shown that these networks influence
attitudes and behavior related to contraceptive use as well as to AIDS. In addition, information on the content of
conversations in these networks is available from standard semi-structured interviews as well as from unusual
ethnographic data.
What neither the survey nor the qualitative data permit, however, is situating informal conversations about AIDS
Abstract: in a wider context. We thus collected new data to address three questions: (1) Is AIDS a major topic of
conversation in a high HIV-prevalence setting where deaths from AIDS are frequent, or are topics, such as food,
income or malaria more frequent? (2) What is the range of topics that are informally but publicly discussed—for
example, do people talk frankly about sexual practices or disapprovingly about people with AIDS? (3) Does the
topic of conversation seem to depend on who is present—e.g. does it change as new people enter the
conversation or others depart?
The paper relies on new data collected in the summer of 2005, using a systematic sampling approach designed
to reduce participant selection bias. Informal topics of conversation and patterns of discussion are collected
through key informants using several new innovative ways (e.g., daily conversation logs, contact diaries and
systematic time sampling).
Session: Personal Network Methods
Keywords: personal network methods , personal networks , affiliation networks , egocentric networks
Accepted: Yes

149 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Elisabeth Gidengil
Authors: elisabeth.gidengil@mcgill.ca
Department of Political Science McGill University Montreal, Canada
Title: Network Diversity and Views about Gay Marriage
Network diversity is often seen as fostering tolerance. However, a recent study by Rochelle Côté and Bonnie
Erickson on the impact of network diversity on attitudes toward ethnic minorities and immigrants found that the
type of diversity is a critical factor: diverse networks only encourage tolerant attitudes if they connect people to
the kinds of people who are likely to be more tolerant. The proposed paper builds on this research by examining
the impact of different types of network diversity on attitudes toward gay marriage. Drawing on the literature on
the gender gap in attitudes towards gays and lesbians, I argue that weak ties with middle- and upper-middle
class women are the most likely to be associated with favorable attitudes towards gay marriage and alternative
lifestyles more generally among men and women alike. The data are taken from the 2000 Canadian Election
Study, a nationwide survey of eligible voters aged 18 and up conducted in connection with the 2000 Canadian
election (www.ces-eec.umontreal.ca/ces.html). The measures of network diversity are derived from a variant of
the position generator originally developed by Lin and Dumin. Respondents were asked whether they knew a
man, a woman, or both, in 15 different occupations ranging from low to high occupational prestige.
Session: Sex, Drugs, and Social Networks
Keywords: attitudes , gay marriage , network diversity
Accepted: Yes

150 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Elisa Giuliani
Authors: giulel@ec.unipi.it
DEA, University of Pisa, Italy SPRU, Sussex University, UK
Firm knowledge base, local embeddedness and external openness:Drivers of innovative performance in wine
On the stream of a long-standing tradition of study in economic geography and regional economics, organisation
theorists and management scholars have in recent times demonstrated an increased interest for industrial
clusters and the innovative performance of their firms (e.g. Porter, 1990; Pouder and St. John, 1996; Bell, 2005;
Tallman et al., 2005; Romanelli and Khessina, 2005). The paper explores the factors that affect the innovative
performance of firms in wine clusters. The analysis is based on a combination of social network analysis and
econometrics. It estimates the effects of four key explanatory variables: firm knowledge base, local
embeddedness, the degree to which a firm bridges structural holes and external openness on the firm innovative
Abstract: performance, using both linear and non linear regression models – General Additive Models (GAM) (Wood,
2000) are used in this second case. The paper makes several contributions. First, it confirms existing studies
about the importance of firms’ knowledge base for innovation. Second, it shows that, in industrial clusters, highly
embedded firms are more likely to be innovative than firms not so positioned. Third, the paper shows that the
search for informational ‘diversity’ by bridging structural holes reduces the chances of being innovative. In
contrast, the search of diverse knowledge outside the boundaries of the cluster is positively related to the
innovative performance of cluster firms. More importantly, this study suggests that firms innovate more through
the absorption of extra-cluster knowledge, than through the embeddedness in local networks.
Session: Inter-Organizational Networks
access to knowledge , embeddedness , information diversity , innovation , inter-organizational , knowledge
transfer , structural holes
Accepted: Yes

151 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Peter Gloor
Authors: MIT, CCI, 3 Cambridge Center, Cambridge, MA 02140, USA
Renaud Richardet
Title: Coolhunting on the Web and in the Blogosphere
This paper introduces a new Web and Blog mining approach, which we call “Web Coolhunting”. We make use of
the fact that the Web has become a mirror of the real world, breaking latest news through active participation of
millions of volunteers on Web sites such as Wikipedia, and political blogs such as dailykos and instapundit.
TeCFlow, a dynamic social network analysis tool, measures popularity and influence of brands and stars over
time by looking at their “social embedding” on the Web. Our approach builds a network map with the linking
structure of a list of Web sites returned by a Google query. Combining multiple datasets, each containing the
linking structure of the Web sites collected through querying Google for the name of a search term (the “star”)
permits to find the most central star in a group of stars by comparing the betweenness scores of the different
stars. By combining the link maps returned by different searches, we can compare different stars, identifying the
ones with the highest betweenness values. These are the most linked, or “talked about” search terms in a given
Web or Blog context.

The same combined query and subsequent evaluation by betweenness also permits to find the most relevant
Web sites discussing related topics. These Web sites also double up as “kingmakers.” For example, pages on
the New York Times, and Wikipedia come up as most relevant without necessarily containing the search terms.
In addition, our dynamic betweenness–based algorithm is much harder to spam that e.g. Google’s page rank,
which is essentially degree-based. We have applied our system to the tracking over time of topics ranging from
political candidates and political issues to music bands and Oscar nominations.
Session: Social Networking Tools
content analysis , data representation , dynamics on networks , internet/hyperlink networks , methodology ,
Accepted: Yes

152 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ertug Gokhan
INSEAD Asia Campus 1 Ayer Rajah Avenue Singapore 138676
Authors: Martin Gargiulo
INSEAD Asia Campus 1 Ayer Rajah Avenue Singapore 138676
Charles Galunic
INSEAD Europe Campus Boulevard de Constance 77305 Fontainebleau Cedex France
Title: Ego network density and reputation: distinguishing local and global effects
This paper explores the moderating effect of network density on reputation. Past research suggests that both
positive and negative reputation should be amplified in denser networks. The probability that information on an
actor’s behavior will reach other actors in the network increases with the density of ties among those actors. The
existing literature, however, has focused on the density of the global network, overlooking the effects that the
local network surrounding an actor (i.e., her ego-network) might have on the diffusion of the actor's reputation.
Consistent with existing research, we argue that the density of an actor’s local network amplifies the reputation
effects of the actor’s behavior at the local level. Yet, local network density can actually mitigate the spread of that
same reputation in the global network. Because actors in a dense local network are less likely to have ties
beyond that local network, information on ego's behavior is less likely to be known to third parties beyond the
local network. Conversely, the behavior of actors with sparse ego networks is less likely to be known at the local
level, but more likely to spread at the global level. A study of a large sample of high-level employees in a global
investment bank provides support for our hypotheses
Session: Intra-Organizational Networks
Keywords: reputation , structural holes , network theory
Accepted: Yes

153 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sandra Gonzalez-Bailon
Nuffield College, University of Oxford, Sociology, Nuffield College, New Road, Oxford, Oxfordshire OX11NF,
United Kingdom
Title: The Structure of the Political Domain: How International Organisations Build Visibility within and Beyond the Web
This paper challenges the view that the Web has democratised access to public opinion, allowing peripheral
organisations to spread a previously silenced voice. While it might be true that the Internet brought a decrease in
the costs associated with political organisation, visibility is still the capital of just a few, exhibiting a rich-get-richer
effect that is not independent from the positions that organisations hold offline. The analysis from which these
conclusions are drawn is based on the citation network that 1001 international organisations build on the Web
(on the .org domain). The assumption in using this data is that the patterns of connectivity between sites (or how
organisations link to other organisations) provide information about their affinity or partnerships and about their
relevance: a high citation is an indication of impact –yet not necessarily of intrinsic worth. Two sets of analyses
are presented: first, information is given on the overall structure of the network, and on the mechanisms that
underlie the formation of ties (considering a number of exogenous attributes like budget, staff and year of
foundation of the organisations); then, the centrality scores of the organisations are correlated with their traffic
ranks (number of individual visits to their sites) and with their media impact (number of times organisations have
been cited by the main international newspapers). The results show that the Web is not the democratic political
marketplace so often assumed to be. The methods used in this paper include OLS, QAP and ERGMs.
Session: Exponential Random Graphs
Keywords: civil society , inter-organizational , mass media , organizations , snowballing , web
Accepted: Yes

154 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Nicholas Gould
University of Glamorgan, Welsh Institute for Health and Social Care, 15 High St, Llantrisant, Pontyclun, Wales
CF72 8BQ, UK
Title: Paradigmatic boundaries in the (social) science of (social) networks: the preliminaries of a Kuhnian investigation
This paper considers paradigmatic boundaries within the science of networks. In respect of the philosophy of
science, the paper has its roots in Wittgenstein’s Tractatus, the concern of Lakatos with pseudoscience and,
more recently, Nancy Cartwright’s contribution describing the Dappled World of scientific endeavour. This core
orientation is substantiated by a comparison between the science of energy during the mid-nineteenth century
and our present day science of networks. The study is Kuhnian in two respects. First, it utilises Kuhn’s prime
concern with the necessity of informing philosophy of science debates by recording the historical context of
development. Historical methodology generates a medium in which, for example, the empiricist debate between
Abstract: Kelvin and Maxwell have a particular salience. The second Kuhnian influence takes the form of Kuhn’s work
specifically on the themes of Kelvin and measurement - “If you cannot measure, your knowledge is meager and
unsatisfactory.” The historical sketch captures the inherent messiness of scientific discovery that ultimately leads
to both intended and unintended consequences of a research programme. This sketch is then used to inform a
discussion of the ontological and methodological boundaries related to the investigation of small world
phenomena. The discussion indicates that what appears to be coherent research – as exemplified by the recent
collection of papers by Newman et al – actually disguises a healthy messiness, often revealed in SOCNET
postings. The paper concludes with a map of paradigmatic boundaries for network science.
Session: Ontology and Philosophy of Networks
Keywords: small world , empiricism , history of science , kuhn , realism
Accepted: Yes

155 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Rick Grannis
University of California, Los Angeles, Department of Sociology, 264 Haines Hall, Box 951551, Los Angeles, CA
90095-1551, USA
Title: Three Degrees of Separation
Neighborhoods have been shown to have the potential for many important effects in the lives of their residents
but many do not actualize this potential. I explore the question of why some neighborhoods are able to form
effective communities while others fail to do so. Using four previously untapped surveys, I show that
neighborhoods must be sufficiently delimited to constrain most of the concatenated paths connecting their
residents, especially their households with children, to about three steps in length. While Milgram and others
have found evidence that about six degrees of separation connect many of us, I find that effective neighborhood
communities need most residents, especially those with children, to be connected by about three degrees of
Session: Neighborhood Communities
Keywords: dynamics on networks , embeddedness , social geography
Accepted: Yes

156 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Harold D. Green, Jr.

Science of Networks in Communities Research Group, National Center for Supercomputing Applications,
University of Illinois at Urbana-Champaign, 1205 West Clark Street, Room 1008, Urbana, IL 61801, USA
Noshir Contractor
Departments of Speech Communication, Psychology, Coordinated Science Laboratory, & Graduate School of
Library & Information Science, Director, Science of Networks in Communities (SONIC), National Center for
Supercomputing Applications, 244 Lincoln Hall, 7
Creating Community: The Role of Arts and Cultural Organizations as Brokers and Catalysts for Increased Civic,
Social, Political, and Cultural Participation among Mexican Immigrants in Chicago
The Mexican Immigrant Assets study builds on recent research on the informal arts as a force for stimulating
civic activism (Wali, Severson and Longoni 2002), and as an important community-building mechanism within
immigrant communities (Moriarty, 2004). The MIA study investigated the roles that cultural, artistic and social
networking practices play in identity formation, community building and the general creative potential of Mexican
immigrants. Our research attempts to penetrate the ‘monolithic’ networks that support transnational communities
to investigate the impact of variation in network members and network structures affect our research themes
(Contractor, Wasserman, & Faust, 2006, Monge & Contractor, 2003).

In this talk, we report on multi-level analyses of the brokering roles played by Mexican-immigrant-serving
organizations in resource exchange networks. We used structural hole and brokerage role detection to
summarize the patterns of resource exchange in this inter-organizational network, finding that organizations often
act as liaisons between different types of organizations, creating pathways among social services, community
centers, arts institutions, businesses, mass media and other organizations important to facilitating access to the
social, political, cultural and economic life of the city for the Mexican immigrant community. Initial models
revealed that variables reflecting the project and funding portfolio of an organization are predictors of an
organization’s liaison activities. Further, multilevel models were used to investigate the effects of organizational
type, organizational demography and resource exchange relationship on the profile of brokerage roles for
respondent organizations.
Session: Ethnic Factors and Immigration
Keywords: immigrants , brokerage , social support , organizations , arts and cultural practices , cultural communities
Accepted: Yes

157 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jeffrey Grierson
Authors: La Trobe University, ARCSHS, 215 Franklin Street, Melbourne, VIC 3000, Australia
Anthony Smith
La Trobe University, ARCSHS, 215 Franklin Street, Melbourne, VIC 3000, Australia
Title: A social network analysis of gay men’s media consumption and organisational affiliations
This paper presents two-mode analyses of media consumption and organisational affiliation in a social network
study of gay men in Melbourne, Australia.

These analyses are drawn from the Victorian Networks Study (Vines), a social network study using an interviewe
administered survey instrument. The project recruited a sample of 213 individuals with a mean age of 37 years.
To assess organisational affiliation the survey asked if respondents were a member of 18 categories of
organisation These categories included gay or lesbian sporting groups, gay or lesbian volunteer organisations,
gay and lesbian religious groups, HIV/AIDS related groups etc. Participants were then invited to nominate up to
two specific organisations within each category. Over two thirds of participants were members of some sort of
formal organisation. Participants who joined organisations were highly likely to join more than one, with 44% of al
participants being members of two or more. To assess media consumption, 36 specific media outlets were listed
and participants asked to rate their consumption on a five point scale.

We present the two-mode analyses of these two domains. For organisational affiliation we assess the structure o
cross-group membership of organisations and examine the significant components that emerge. For media
consumption we assess the structural relationship in terms of both co-occurrence and strength of tie (as
measured by consumption rating). For each analysis we discuss how the findings were workshopped with HIV
prevention organisations and how the structural analysis was able to enhance health promotion practice.
Session: social networks and health
Keywords: 2-mode , affiliation networks , community structure , hiv , peer networks
Accepted: Yes

158 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dave Griffiths
Authors: davegriffiths77@supanet.com
University of Manchester, Sociology, Flat 2, 70 Park Road West, Claughton, Wirral CH43 8SF, England
Title: The Outside Interests of the UK’s Cultural Agencies
This paper focuses on UK non-departmental public bodies and the range of organisations and institutions
represented by those individuals sitting in these public boardrooms. Biographical data has been collected on all
board members of the 44 executive agencies of the Department of Culture, Media and Sport. This data originates
from the organisation’s annual reports, registers of members’ interests and biographical directories. This
information details connections which are economic (interlocking directorships, employers and academic ties),
social (charity trusteeships, positions within non-profit organisations and memberships of professional bodies and
gentlemen’s clubs) or quango-related (advisory roles, former positions and agencies outside the department).
Social network analysis has been used to identify which organisations and types of connections carry the
greatest social capital within this network. Similarly the data has been analysed to understand how this social
capital is used by both the cultural agencies and the outside organisations. This study also uses blockmodeling to
examine whether there is evidence of these individuals forming an elite, and whether there is an ‘inner core’ of
agencies at the heart of this network. This study aims to understand the mutual influences and settings which
directors of cultural agencies share outside of the boardroom to help understand which interests are being
represented within them.
Session: Politics and Interlocking Directorships
Keywords: interlocking directors , political networks , political culture , social capital
Accepted: Yes

159 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Francesca Grippa
Authors: University of Lecce- Italy, eBusiness Management Section -S.S. ISUFI, via per Monteroni, Lecce 73100, Italy
Peter Gloor
MIT Sloan Center for Collective Intelligence Cambridge (MA) USA
Title: Selective Memories Favor Influentials
Social scientists have long been interested in understanding what makes some respondents more accurate
observers of their own social networking behavior than others in recalling and reporting social network’s structure
(Bernard et al. 1984, Freeman et al, 1987, Casciaro, 1998).
This study contributes to social network research on interpersonal perception accuracy, by focusing on the
relationship between actors’ centrality and their ability to report accurately their own social interactions. We used
network measures like actors’ betweenness centrality and degree centrality to identify the most prominent
This research was conducted within a graduate school of management that both educates students and
undertakes research projects. Through seven web surveys, we gathered data on the frequency of interaction
Abstract: among members connected through telephone, chat, and face-to-face conversations. We conducted focused
interviews with three senior managers to classify all network members according to three factors: trustworthiness
prestige, and contribution.
We applied an “internal measure of accuracy” by measuring the correlation between ego-perception and
alter-perception. We found that actors involved on the same interaction attributed different values to the same
interaction (Riji ≠ Rjij), which is referred to as a “non-reciprocity” type of misalignment (Krackhardt, 1987).
We found a positive correlation between actors’ centrality and their centrality as assessed by senior
management, and a negative correlation between actors’ centrality and their accuracy in recalling interactions.
These asymmetries suggest that underreporting social interactions may represent a third way of measuring the
importance of members and finding the most influential ones.
Session: Communication Networks
Keywords: centrality , inter-individual , intraorganizational networks , leadership and networks , network surveys
Accepted: Yes

160 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Piotr Gronek
AGH-UST Faculty of Physics and Applied Computer Science, Cracow, POLAND
Authors: Malgorzata Krawczyk
AGH-UST Faculty of Physics and Applied Computer Science, Cracow, POLAND
Antoni Dydejczyk
AGH-UST Faculty of Physics and Applied Computer Science, Cracow, POLAND
Title: Structure of a local e-mail network: students, staff and noise
We investigate the cluster structure of e-mail addresses at the Faculty of Physics and Applied Computer
Science, AGH-UST in Cracow. The input data is the number of messages, collected between September 2006
and January 2007 at our local server. This directed network is the basis for the weighted connectivity matrix and
the modularity index Q, as defined by M.E.J.Newman (PR E 70 (2004) 056131). Calculations of Q allow to
identify clusters of e-mail writers, the relative strength of correlations inside and between observed clusters. The
structure is obtained separately for students and staff. The results reveal small groups of continuously
Abstract: cooperating scientists, embedded in the structure of departments. Also, some hubs can be identified in the
network. These addresses are used for administrative purposes, to send advertisements and messages to the
staff and to students. During a recent reorganization, some departments were grouped in larger ones. Careful
analysis allows to capture links between departments in the old structure, which could be used to predict the
emerging new pattern. Investigating the network of students and staff together, we could trace communication
within the internal specializations. However, this structure is more noisy, and the network limited to students is
very sparse.
Session: Computer Networks as Social Networks
Keywords: e-mail database , clustering
Accepted: Yes

161 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Michel Grossetti
Authors: Michel.Grossetti@univ-tlse2.fr
CNRS, UTM 5 Allees A Machado, Toulouse, Cedex 31058, France
Title: Social relations and mediation resources in the creation of innovative companies.
The creation of a company is a complex process that involves many actors, and especially different levels of
action. In this process, the founder or founders, considered as individual actors, give rise to a collective actor, an
organization, by mobilizing resources and relying both on individuals and on existing organizations, sometimes in
a more global milieu or an established market. We used crossed interviews (minimum 2 per case) and the
analysis of web information for reconstructing the processes of creation of 50 innovative companies in the south
west of France (in Toulouse, in some towns one hour drive from Toulouse, plus some cases in Bordeaux and
Montpellier). These companies were selected for their innovative character, attested by the fact that they
received innovation subsidies or are housed in a nursery. They are located in the sectors of IT, biotechnologies,
chemistry, and mechanics. For each case, we wrote a summary of the story that we sent to the interviewed
persons who were asked to propose corrections.

In this communication, I will especially focus to situations of access to resources. Mobilization of a resource is a
sequence of actions in which one of the creators mobilizes or receives a resource he does not have. For
example, consultation with an attorney to draw up the articles of association is a sequence that involves one or
more creators, the attorney, and the advice he dispenses. The attorney may be a relation of one of the creators,
or a relation of a relation, in which case we consider that mobilization of the resource is achieved through social
relations, or he may have been selected from a directory, in which case we consider that this is a mediating
resource that allowed access to the resource. Hence, for example in case n. 107, the sentence (made
anonymous here) “The articles of association are drawn up voluntarily by an auditor whom [the creator] met while
dancing” makes it possible to code that the resource is a counsel, that it was obtained through mobilization of a
relation (by a relational chain of length 1), a relation established during leisure time. The data analysis is still
ongoing. For the first 40 companies we studied, the number of situations of resource mobilization averaged 15.3
per history (the minimum is 2 and the maximum 34).

The communication will present, for different stages of the creation process, results concerning the proportion of
resources mobilization by the way of social relations chains, the length of these chains, the kinds of relations
which are mobilized, the kinds of mediations resources that are used, the links between the kinds of resources
and the way founders access to them.
Session: Networks, Economics, and Markets
Keywords: economic networks , qualitative network , social capital , mediation resources , embeddedness
Accepted: Yes

162 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Wojciech Gryc
University of Toronto, Centre for Urban and Community Studies, Suite 400, 455 Spadina Ave. Toronto, ON
Authors: M5S2G8, Canada
Bernie Hogan
University of Toronto, Sociology, Department of Sociology, 725 Spadina Ave., Toronto, ON M5S2J4, Canada
Title: A new measure of social distance
Homophily is a prominent and consistent measure in social networks. It has been found so frequently in social
networks that McPherson and Smith-Lovin (2001) suggested it has 'law-like' properties. That said, measures of
homophily rarely take into account the overall connectivity of the graph. Common measures either consider the
expected proportions of people in separate groups, or count direct edges between individuals of like and unlike
type. This can lead to unsatisfactory conclusions about the make-up of a network. For example, a network
(visualized as a bowtie) where a central person is tied to two male-female couples would appear similar to a
network where an individual is tied to two male friends and two female friends, despite the fact that these
networks exhibit clearly differing organizing principles.
We seek to resolve this discrepancy (and other similar discrepancies) through the use of a novel measure of
'social distance' D. This measure compares the average closeness between different categories of individuals
within a mixing matrix. We believe this measure will be a fruitful way to assess whether individuals are clustered
together in subgraphs of like type or distributed throughout the entire network. We test the value of this measure
over a series of graphs, both random and empirical. We find that compared to similar measures, particularly Mark
Newman's measure of assortativity, ours is more stable across networks of varying density and takes into
account indirect ties. Additionally, our measure takes into account unconnected networks, which – for reasons we
outline in the paper – makes it a novel and useful measure for ego-centered networks.
We conclude our paper by illustrating how this measure can be used in the field. We give the example of social
distance between schoolchildren of different races and genders from the AddHealth dataset.
Session: Algorithms and Analytic Methods
Keywords: algorithm , homophily , methodology , assortativity , egocentric networks
Accepted: Yes

163 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Christian Gulas
FAS.research GmbH, Vienna, Austria
Anna Schreuer
Authors: IFZ - Inter-University Research Centre for Technology, Work and Culture
Max Ruhri
FAS.research GmbH, Vienna, Austria
Harald Katzmair
FAS.research GmbH, Vienna, Austria
Title: The Gender of Basic Research. Comparing Knowledge Networks of Women and Men
The presentation describes the gender-specific patterns of knowledge networks in basic research. Knowledge
networks are operationalized as sets of sciences that co-occur within research projects funded by the Austrian
Science Fund (FWF) between 1992 and 2006. Sciences are classified according to the OECD-taxonomy of
scientific disciplines (4-digit codes).
Scientific disciplines that are linked through projects generate networks of knowledge production. The output of
knowledge produced by one scientific discipline constitutes an input for all the sciences it is connected with. The
output of those sciences then further constitutes an input for others, and so on. This produces emergent patterns
of knowledge production and exchange, autocatalytic sets and clusters, which are found to differ in the networks
Abstract: of women and men.
One important finding of the study is the gender difference with regards to brokering sciences. For both men and
women the network displays a grand divide between the natural sciences and the social sciences/humanities;
however it is bridged by different broker-sciences in the networks of women and men.
Furthermore, with regards to gender, the study found the network structures to bear greater resemblances in the
natural sciences and human medicine than in the social sciences or in humanities.
The research systematically analyzes the similarities and differences with various mathematical and visualization
techniques. One focus lies in the search of autocatalytic sets - highly dynamic network areas of cyclic exchange
between sciences of basic research.
Session: Knowledge Networks
Keywords: knowledge networks , knowledge production , innovation
Accepted: Yes

164 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ido Guy
Authors: ido@il.ibm.com
IBM, Research, University Campus, Mount Carmel, Haifa, Israel
Title: SONAR - Social Networking Architecture
Social software is proliferating in the Web 2.0 era: from blogs and Wikis through recommender systems (such as
Amazon, eBay, IMDB) to social tagging (such as Del.icio.us) and personal networking systems (such as Orkut
and LinkedIn). Yet, social networks can also be derived by other means of collaboration, such as project
membership, document co- authorship, conference meetings etc.
SONAR is an API and a data format for sharing social-network data and aggregating it across applications to
show who is related to whom and how. Where possible, open standards such as Atom and JSON are used, and
service APIs follow the REST design pattern. Clients of SONAR could vary from expertise miners to
visualizations to user interface widgets. SONAR will answer questions like, "who does this person work with
Abstract: most?", "who has tagged this person?", "what are all the artifacts produced in division 22 this month", "who
should I cc on my email with subject 'SONAR design goals'?"
SONAR clients can use the SONAR API to access data from a single provider that implements the API.
However, the more compelling case is where an intermediate component, an aggregator, is used by clients, with
the very same SONAR API, to consolidate the data from different providers. This way, one can choose multiple
data providers and assign an appropriate weight to each of them.
We demonstrate the potential benefit of using the SONAR API over an extendible instant messaging and
collaboration client. Our extension uses eight different data providers to show valuable social network
information, such as related people, connection points, and people in common.
Session: Social Networking Tools
Keywords: access to knowledge , collaboration , data collection , standardization , social networking api
Accepted: Yes

165 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dirk Günther
Authors: Institute of Environmental Systems Research Barbarastr. 12 University of Osnabrück D-49069 Osnabrück
Jens Newig
Institute of Environmental Systems Research Barbarastr. 12 University of Osnabrück D-49069 Osnabrück
Title: Using Social Network Analysis to explain Social Learning in Policy Networks
In the face of apparent failures to govern complex societal problems by central state top-down oriented
policy-making, social or collective learning in network governance becomes an es-sential feature. Social learning
not only refers to cognitive changes in individuals within the network, serving as a structural framework, but is
moreover understood as a process in which individual changes in cognition lead to modifications in the collective
network structure.
Although SNA provides powerful tools for recognizing patterns in network structures, this area of research seems
up to now largely unconnected to the literature on social and collective learning. This paper aims to investigate
- network characteristics that foster social learning in policy networks and
- how social learning becomes manifest in changed network structures.
Arguably, certain network structures are more apt to foster collective learning than others. Partly, this depends on
the defition of social learning. When focussing on efficient informa-tion transfer, highly centralized networks
appear advantageous; network density and low cen-tralization become an essential features when true
deliberation is aimed at. Then again, net-work resilience (redundancy) may be important for long-term effective
social learning.
On the other hand, we expect network structures to change as social learning takes place. New relations are
established while others are dissolved; moreover, competencies, roles and tasks of network actors can be
expected to change and develop in the course of social learning. Thus, network structure can be conceived both
as independent and as dependent variable.
Session: Policy Networks and Governance
policy network , network structure , knowledge transfer , inter-organizational learning , civil society ,
governance , network dynamics , social learning
Accepted: Yes

166 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Steven Haas
School of Social and Family Dynamics Arizona State University Tempe, AZ
Authors: David Schaefer
School of Social and Family Dynamics Arizona State University Tempe, AZ
Olga Kornienko
School of Social and Family Dynamics Arizona State University Tempe, AZ
Title: A Friend in Need: Health and the Structure of Adolescent Social Networks
Researchers from the social and medical sciences have long investigated the protective role of social support
and the variety of resources derived from social networks in maintaining health. Such research tends to take the
existence of social networks, and the support they provide, as a fact of social life that is a priori exogenous to
health status. As such, scant research has examined the converse, how health may structure the creation and
maintenance of social relationships in the first place and almost none has used a social network perspective to
investigate how health may influence the structural properties of peer networks. This paper begins to fill this gap
by providing a theoretical and empirical elaboration of the ways in which health may shape various aspects of
Abstract: adolescent peer networks. We hypothesize that adolescents in poor health are likely to direct limited personal
resources to fostering social networks with structural properties that maximize the provision of social support.
Specifically, we hypothesize that the social networks of adolescents in poor health will be characterized by strong
ties within a smaller, higher-density, local network, rather than weaker ties in a larger, more diffuse network.
Ultimately, the resulting peer relationships reflect both the limitations imposed by their illness and the constraints
of normative adolescent network processes. Utilizing the National Longitudinal Study of Adolescent Health we
provide an empirical assessment of the impact of various dimensions of health on the structure of adolescent
peer networks.
Session: Adolescent Friendship Networks
Keywords: egocentric networks , health , social networks , friendship networks
Accepted: Yes

167 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Roger Haeussling
Authors: r.haeussling@gmx.de
University Karlsruhe Germany
Title: Interactions, Interventions, Semantics, and Emotions: A Multi Level Network Approach
We would like to introduce a network theoretical and -analytical four-level concept for the indication of social
interactions. It assumes the following levels: the context of interactions, the network of the interlaced interactions
itself with its own dynamics and (positional) constellations. The interventions of the actors involved are subtly
distinguished from the interactions. They are seen as an attempt to get involved in the ongoing process with
one’s own personal contribution.
In doing so, an analytical distinction can be made between two interventional levels: the active, physical
interventions, e.g. a disclosure, or a certain way of acting and receptive interventions in form of interpreting a
situation, the ongoing interactions by indicating other actors and their contributions. The revelation of emotions
as the fourth level, which is necessary to transform a relation of dependency to a relationship, like a friendship, is
dealt with as a special form of intervention.
Every level has its own method of empirical registration. For the network level of interactions one can use for
example a quantitative network-analysis. For the two levels of interventions the qualitative analysis of networks
plays a very prominent role. This contribution plans to illustrate this analysis more closely. Here, an emphasis
has to be made on the fact that this analysis is the most qualified for triangulating its results with the results of the
qualitative network analysis and thus to amend the topographical sight of the network by an actor-based narrative
prospect on the network, in which the actors are embedded.
Session: Qualitative Network Research: Methodological and Theoretical Issues
Keywords: methodology , interaction , multi-level networks , network theory , emotions , intervention
Accepted: Yes

168 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Valerie Haines
University of Calgary, Department of Sociology, Calgary, Alberta T2N 1N4, Canada
Authors: John Beggs
Louisiana State University Department of Sociology Batton Rouge, USA
Jeanne Hurlbert
Louisiana State University Department of Sociology Batton Rouge, USA
Title: Neighborhood disadvantage,network social capiptal and depression
The focus of research on neighborhood effects has shifted from demonstrating that neighborhoods matter for the
health of individuals to identifying mechanisms that explain why they matter. We explore the role that network
social capital, understood as network structures and resources embedded in these structures, plays in mediating
the relationship between neighborhood disadvantage and depression. We find that living in a disadvantaged
neighborhood increases depression, over and above the characteristics of residents. We also find that network
social capital mediates the association between neighborhood disadvantage and depression, and that this role
can be more fully understood only by considering both the structure and resource elements of network social
capital. Because residents’ perceptions of neighborhood disorder have been shown to mediate the association
between neighborhood disadvantage and depression, we examine whether perceived neighborhood disorder
accounts for the effects of network social capital on depression. Our finding that it does not highlights the
importance of paying close attention to characteristics of neighborhoods and characteristics of networks in
research on how neighborhood disadvantage affects health.
Session: social networks and health
Keywords: health , neighbourhood and health , social capital and health
Accepted: Yes

169 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dan Halgin
Boston College, Organization Studies, Fulton Hall Room 430, 140 Commonwealth Avenue, Chestnut Hill, MA
Authors: 02467-3808, USA
Steve Borgatti
Boston College, Organization Studies, Fulton Hall Room 430, 140 Commonwealth Avenue, Chestnut Hill, MA
02467-3808, USA
Title: Thank You For Dissing Me: The Phenomenon of Status Flows
In this paper we analyze status contagion among actors in a cultural industry. We begin with a discussion of the
consequences of having relationships with high status actors, drawing upon social capital literature, affiliation
literature, and status attainment literature. We contend that putatively negative interactions with high status
actors can confer status to lower status actors much like positive interactions. Specifically, we analyze the
Abstract: phenomenon of “diss songs” among rappers in hip hop music. A diss song is a song in which the performing
rapper makes derogatory comments about another rapper. Diss songs are recorded and released on mixtape
albums and often receive radio play. We analyze 870 directed disses among 218 rappers over 20 points in time.
Findings suggest that being publicly “dissed” by a high status actor increases a rapper’s status and attracts
additional disses, creating a virtuous cycle of success.
Session: Social Capital
Keywords: dynamics on networks , social capital , status flows , cultural industries , contagion , negative ties
Accepted: Yes

170 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Mark Handcock
University of Washington, Department of Statistics, Box 354320, Seattle, WA 98195-4320, USA
Martina Morris
Authors: University of Washington, Department of Statistics, Box 354320, Seattle, WA 98195-4320, USA
David Hunter
Pennsylvania State University, Department of Statistics, University Park, PA, 16802, USA
Pavel Krivitsky
University of Washington, Department of Statistics, Box 354320, Seattle, WA 98195-4320, USA
Title: Using ERGM models for dynamic network simulations with vital dynamics
To simulate a dynamic network, it is necessary to address three issues: the structural regularities, the dynamics
of partnership formation and dissolution, and nodal vital dynamics (e.g. births and deaths). In this paper we
demonstrate a model-based approach for simulating dynamic networks with arbitrary vital dynamics and structure
specified by an ERGM. The model is implemented with a multi-step algorithm: (1) estimate the structural
parameters of an initial network conditional on population composition (as described in the talk by David Hunter)
Abstract: and use these to estimate the corresponding natural ERGM parameters; (2) dissolve partnerships using a
survival model, and apply vital dynamics to the nodes; (3) form new ties using a conditional ERGM with structural
parameters and updated population composition parameters. We show that using that this method produces a
network that is stationary in the structural parameters of interest. We apply the method to data on Ugandan
sexual networks, and show how the method can be used to simulate the spread of HIV through a dynamically
changing network.
Session: Friendship networks
Keywords: ergm , dynamics , egocentric data , simulation , infectious disease
Accepted: Yes

171 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Steven Harper
University of Illinois at Urbana-Champaign National Center for Supercomputer Applications Urbana, Illinois, USA
Carter Butts
Authors: University of California at Irvine Department of Sociology Irvine, California, USA
Ryan Acton
University of California at Irvine Department of Sociology Irvine, California, USA
Noshir Contractor
University of Illinois at Urbana-Champaign Department of Speech Communication Urbana, Illinois, USA
Mapping and Modeling multidimensional emergent multi-organizational networks (EMONs) in the aftermath of
This study reports on a project that advances our understanding of interorganizational coordination in disaster
response by estimating and analyzing the emergent multi-organization networks (EMONs) involved in the
response to Hurricane Katrina. By being able to capture EMONs while they are being formed and evolving, it may
be possible to diagnose those networks for points of potential communication failure, and in turn adjust resource
allocation to prevent failure. Using novel computational methods (specifically text mining techniques), the
research captures, validates, and integrates data from web sites (specifically “situation reports”) to produce
estimates of inter-organizational interaction over time. Additionally, manual methods were used to estimate the
networks present and these results were compared with the automated computational methods. The networks
captured from this data mining are analyzed to determine their suitability as representative models of the EMONs
formed during the disaster. Further, these estimated networks were used to test multi-theoretical multilevel
(MTML) models to explain organization’s motivations to create, maintain, dissolve, and reconstitute the
inter-organizational links. The models were tested using exponential random graph modeling (ERGM)
techniques. Preliminary results using the automated coding methods are consistent with a MTML contextual
model of organizations mobilizing to create networks.
Session: Inter-Organizational Networks
emergent structure , inter-organizational networks , network analysis of text , network evolution , emergency
Accepted: Yes

172 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Andreas Harrer
University of Duisburg-Essen, Department of Computer Sciemce and Applied Cognitive Science, Duisburg,
Sam Zeini
University of Duisburg-Essen, Department of Computer Sciemce and Applied Cognitive Science, Duisburg,
Sabrina Ziebarth
University of Duisburg-Essen, Department of Computer Sciemce and Applied Cognitive Science, Duisburg,
Title: Visualization of the Dynamics of Computer-mediated Community Networks
In this paper we will demonstrate the potential of processing and visualizing the dynamics of computer-mediated
communities by means of Social Network Analysis. According to the fact that computer-mediated community
systems are manifested also as structured data, we use data structures like e-mail, discussion boards, and
bibliography sources for an automatic transformation into social network data formats. Currently our developed
converters DMD (Data Multiplexer Demultiplexer) and DDC (Dynamic Data Converter) support GraphML,
UCINET, and Pajek formats besides our own data formats which are used for real-time analysis of CSCL
(Computer Supported Collaborative Learning) activities. In the case of communication data our converters utilize
conversation graphs reflecting aspects of speech act and conversational theory to produce directed graphs in the
Abstract: cases where one-mode person networks are desired.

The paper will demonstrate a 3-dimensional visualization of an author community based on BibTex bibliography
data converted into GraphML. Based on this dataset we visualize publications network with a tool called Weaver,
which is developed in our research group. According to Lothar Krempel’s algorithm, Weaver uses the first two
dimensions to embed the network structure within a common solution space. The third dimension is used for
representing the time axis and thus the dynamics of co-authorship relations.
Concluding we aim to discuss potential issues and problems of our approach and the possibilities especially
concerning the appropriate visualization and segmentation of long term communications, such as mailing lists.
Session: Visualization
visualization , dynamics on networks , network evolution , paper co-atuhorship network , social network
analysis tools , 3-dimensional visualisation
Accepted: Yes

173 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Nicholas Harrigan
Australian National University & Oxford University, Department of Politics and International Relations, Manor
Road, Oxford, OX1 3UQ
The Inner Circle Revisited: Using exponential random graph models (ERGM) to study the political activity of
corporate elites.
This paper examines the interlocking directors networks of executive and non-executive directors of the largest
250 corporations in Australia. Using exponential random graph models and a detailed dataset of the economic,
social and political characteristics of these corporations and their directors, the paper asks "What types of
corporations tend to be tied to each other by interlocking directors?" The study found strong evidence for a
version of the 'inner circle' thesis (Useem, 1984), with an active core of highly networked directors who span
multiple spheres of the economic and political life of business. In particular, the study found that corporations that
were highly interlocked tended to have four major characteristics: these corporations tended to be (1) large
corporations (as measured by revenue), (2) interlocked with government enterprises, (3) interlocked with the
executive committees of the major business associations, and (4) directed by prominent or prestigious directors
(as measured by listing in Who's Who).
Session: Politics and Interlocking Directorates
Keywords: political networks , ergm (p*) , interlocking directors , business networks
Accepted: Yes

174 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jenine K. Harris
Saint Louis University
Authors: Douglas A. Luke
Saint Louis University
Rachael Zuckerman
Brandeis University
From Discovery to Dissemination: Using Citation Network Analysis to Examine Connections Between two
Branches of Secondhand Smoke Research
Health related scientific discovery, such as linking lung cancer with smoking, leads to interventions designed to
reduce risk and improve health. Over the past few decades the scope of research into the health effects of
secondhand smoke (SHS) has grown immensely. The resulting knowledge has informed interventions ranging
from educating new parents about smoking in the home to developing statewide clean indoor air legislation. The
goal of this project was to examine connections between the evolution of knowledge about the health effects of
SHS and the development and implementation of related interventions. In order to examine how these two
Abstract: branches of SHS research were connected we identified 1880 empirical articles published between 1965 and
2005. From this set of articles we developed two citation networks: one including 1491 articles about the health
effects of SHS and the other including 389 articles about SHS related interventions. We examined the ties
between these two citation networks and their main paths. Based on this examination we present methods for
understanding the relationships between two citation networks embedded within a single scientific area. In
addition, we present information allowing a better understanding of the link between health related scientific
discovery and public health interventions.
Session: Public Health Networks
Keywords: citation network , health , smoking , paper citation network
Accepted: Yes

175 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Koichi Hasegawa
Authors: k-hase@sal.tohoku.ac.jp
Tohoku University, Department of Sociology, Sendai, Japan
Title: Collaborating Environmental Network on Global Climate Change Issue in Japan
Two years ago, Feb., 2005, the Kyoto Protocol was finally enacted. The next year 2008 is the start of the first
engagement until 2012 to decrease the emission level of green gas. The Kyoto Protocol is the only authorized
UN document which title includes the name of a Japanese city. But I regret to say that, even in Japan our
accomplishment of global warming protection in these 9 years is not so good, rather getting much more serious
year by year. Already the emission level in Japan goes up to 8 % and more. Japanese government promised a
reduction of 6% compared to 1990 level. So we must decrease totally 14% of green gas emission until the final
target year 2012. It looks like hardly to reach. Such Japanese experience shows the difficulty of global warming

My focus of this paper is in the policy network and collaboration formation among government sector, private
sector, and citizens’ sector on global climate issue. I would like to call this network “collaborating environmental
network” . I describe an example of such “collaborating environmental network” in recent Japan.

In 1998, right after the Kyoto conference, the Japanese government passed a new law to promote global
warming protection. Under this law, each prefectural government is encouraged to nominate a prefectural center
of promoting global warming protection. Currently, 90% of whole 47 prefectures have such centers except Tokyo
Metropolis and other conservative 4 prefectures. Many centers are a kind of affiliated organization of prefectural
government, Gaikaku Dantai, but some are independent from local government. In the case of my prefecture,
Miyagi prefecture, this center is completely independent and was founded by a local environmental NGO,
MELON, Miyagi Environment Life Out-reach Network in 2000. These 42 centers consist a nationwide liaison

Japan has the following major three levels of networks on the global climate issue; the first two are official and
the third is voluntary.
a) National governmental level network (vertical) : Ministry of Environment---- Prefectural government office of
b) Prefectural government level network (vertical): Prefectural government office of environment---- Prefectural
center of promoting global warming protection----- Action initiator of promoting global warming protection
c) Nationwide liaison network (horizontal) : Each prefectural center of promoting global warming protection.

My paper focus on the basic structure, functioning and sociological characteristics of these networks.
Session: Policy Networks on Climate Change
Keywords: horizontal/vertical networks
Accepted: Yes

176 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sebastian Haunss
Authors: haunss@sozialwiss.uni-hamburg.de
University of Hamburg, Department of Social Sciences, Institute of Political Science
Title: Collective Action Networks in Two European IPR-Conflicts
In our paper we will present preliminary findings from our research project on conflicts about intellectual property
rights in Europe.
In a global, informational economy access to knowledge, production of knowledge, and private or collective
acquisition and control of knowledge are increasingly attracting political attention. The dominant »maximalist
rights culture« of international organisations and agreements (GATS, WIPO, TRIPS) leads to a mode of
regulation of intellectual property claims that expands the protection of intellectual property in scope and depth.
But with the increasing economical importance of intellectual property claims, political conflicts are emerging,
contesting the structure, scope, and boundaries of existing legal frameworks of intellectual property rights (IPR).
In these emerging conflicts the dominant maximalist position is fundamentally questioned.
In our research project we focus on conflicts about two EU directives – the EU directive on the enforcement of
Abstract: intellectual property rights (IPR Enforcement Directive) and EU directive on the patentability of
computer-implemented inventions (Softwarepatent Directive). The study aims to explain why in the first case the
decision making process followed the dominant maximalist rights culture, while in the second case the arguments
of the proponents of alternative regulation modes have been heard.
Starting from the assumption that, as a result of parallel decision-making structures and nearly identical
time-frames, the differences between these decisions can not be explained at the institutional level, the research
project searches for answers on three other levels: actor constellation, framing/argumentation, and forms of
In our contribution to the conference we will present the preliminary results of a network analysis of the actor
networks of both conflicts, and show how the different structure of theses networks may explain the surprising
Session: Civil Society Networks
policy network , globalization , social movements , collective action , intellectual property , access to
Accepted: Yes

177 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Peter Hedström
Nuffield College, Oxford University, Oxford, UK.
Authors: Monica K Nordvik
Department of Sociology, Stockholm University, Stockholm, Sweden.
Ka-Yuet Liu
Nuffield College, Oxford University, Oxford, UK.
Title: Interaction domains and suicides
This paper examines how suicides among those with whom an individual interacts influence his or her own risk of
committing suicide and the ways in which such effects are conditioned by the interaction domains of the
individuals. We focus on two types of domains – the family and the workplace – and we analyze how suicides
within these domains influence the suicide risks of others within the same domain. Data from a unique Swedish
database of all individuals who ever lived in the greater Stockholm metropolitan area during the 1990s are used.
Our results show that the risk of committing suicide is indeed affected by the suicides of others, and that the
magnitude of this effect – the individual as well as the public health effect – differs greatly depending upon the
interaction domain within which a suicide takes place.
Session: Social Networks and Behavioral Change
beliefs , contagion , health , interaction , kinship network , social influence , social relationships , sphere of
Accepted: Yes

178 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Eelke Heemskerk
University of Amsterdam Department of Political Science Oudezijds Achterburgwal 237 1012 DL Amsterdam The
Title: Corporate-Government Networks in the Netherlands: Interlocks between business and politics
This paper investigates the shifting relationships between business and politics over the last quarter of the 20th
century in a coordinated market economy: the Netherlands. Overlapping relationships of interlocking directorates
on the one hand, and governmental advisory boards on the other, as well as political positions of (former)
Abstract: corporate directors tie together a web of corporate-government relationships. As this paper argues, this network
had its heydays during the mid-seventies. Ever since, the gap between business and politics has been growing.
This is understood as a dynamic of competing elites, under the pressures of political and economic
Session: Politics and Interlocking Directorates
Keywords: political networks , governance , corporate social capital , interlocking directors
Accepted: Yes

179 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Margaret Hellard
Authors: hellard@burnet.edu.au
Centre for Epidemiology and Population Health Research Burnet Institute, Melbourne, Australia
Title: Using a social network approach to better understand hepatitis C immunovirology in injecting drug users.
Our first study of the hepatitis C virus (HCV) in the social networks of injecting drug users (IDUs), Networks I
(N1), produced several important insights, most notably that some IDUs remain HCV-free despite injecting and
sharing needles with infectious partners suggesting the existence of protective immunity to HCV. This and other
finding led to the establishment of our second HCV networks study of IDUs, Networks II (N2).

N2 is a two year longitudinal network study that aims to identify the behavioural, virological and immunological
factors affecting the dynamics of HCV infection, re-infection or super-infection and to determine the behavioural,
virological and immunological factors affecting HCV free survival in a cohort of IDUs. Over 300 participants have
been recruited since the study commenced in 2005. As well as completing a questionnaire about sexual, injecting
and other risk behaviour, network data, and socio-demographics, participants provide a 50ml blood sample which
is tested for HCV RNA, as well as HCV core gene sequences. The peripheral blood mononuclear cells are
isolated and HLA typed then tested for HCV specific CD8+ T cell responses.

Currently the IDU network is used somewhat pragmatically to inform our understanding of HCV incidence and
transmission, be they behavioural or immunological, to inform future HCV vaccine development. At the
completion of the study the network will be interrogated more fully so we better understand how the network as a
whole influences HCV transmission and the implications of this if a HCV vaccine is to be successfully delivered to
this population.
Session: Infectious Diseases and Social Networks
Keywords: drug use , hepatitis c , social network
Accepted: Yes

180 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Iina Hellsten
Authors: Royal Dutch Academy of Arts and Sciences, Amsterdam, The Netherlands
Rudy Prabowo
Statistical cybermetrics, University of Wolverhampton, the UK
Title: Networks of communication on the Web: The bird flu hype
The paper discusses the dynamics of public communication networks during the bird flu hype in 2005-2006. In
particular, the focus is on the spreading of the frame of bird flu as a potential influenza pandemic -- comparable
to the Spanish Flu in 1917-1918 that killed millions of humans. The pandemic frame became dominant in public
debate on bird flu in October 2005 when scientific research pointed into structural similarities between the bird flu
and the Spanish Flu viruses. The amount of published texts on bird flu in newspapers, blogs and (popular)
medical journals suddenly multiplied in between September and October 2005.

The aim of the paper is to develop new ways of analyzing the spread of key frames and ideas in and across
Web-based communication networks. Instead of focusing on hyperlinks, the paper considers the pandemics
frame as a link, and published texts as nodes in the network. The pandemics frame is followed in and across the
published texts, news items as well as discussion postings, in on-line news media (BBC online), blog postings
(BlogPulse), and on-line news services (Google!News) from October 2005 to June 2006. The results of the study
show that connections between these domains amplified when the pandemics frame became popular, resulting
both in intensification and fragmentation of the debate. The study takes part in recent research on on-line
communication networks as a special case of social network analysis, and opens up new avenues for research
concerned with cross-domain communications.
Session: Network Dynamics
Keywords: dynamics on networks , framing , internet/hyperlink networks , network dynamics , network coevolution
Accepted: Yes

181 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Marina Hennig
Humboldt Universität zu Berlin, Institut für Sozialwissenschaften, Universitätsstrasse 3b, Berlin, - 10017,
Title: To which aim are existing personal networks generated?
With regard to capturing social relations most approaches up to now (Lin, Granovetter, Fischer, Wellman) have
argued from a perspective of Network Theory. They presume either functions or contexts to be the motives or
basis for the constitution of networks. As an example, the distinction between strong and weak ties is based upon
networks associated with different occupational fields. The same applies to Lin’s position generator, which
situates network agents according to the dimensions of their occupation (job prestige).
Barry Wellman’s community approach, however, focuses on the significance of social context as a central theme.
Claude Fischer, too, attempts to constitute networks by means of the social context, but fails to base these on
sufficient theoretical differentiation.
The issue neglected in the aforementioned approaches is that individuals act within institutionally constituted
Abstract: structures. A child, for instance, acts in certain contexts that vary depending on its age and which serve to
constitute the social network of the child. Thus, the previous approaches may fail to detect empirically very
different or further reaching network constellations due to insufficient theoretical conceptions of the institutional
constitution of structures and action space. That would mean that because of a theoretical position we fail to
notice something which may in actuality be present.
Hence, my paper addresses the questions of how institutionally constituted structures can be translated into
generators which capture social relations in personal networks and how activities can be reconstructed through
social relations. For this purpose, a study conducted in three German cities to survey the networks of 1953
families with children under the age of 18 in their household is going to illustrate how we can succeed in
collecting data on such institutionally constituted structures and, thus, generate corresponding networks.
Session: egocentric network
Keywords: egocentric networks , family , personal networks , methodology
Accepted: Yes

182 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Julia (Julie) Hersberger

University of North Carolina at Greensboro, Library and Information Studies, P.O. Box 26170, Curry 349D, 1000
Spring Garden Street, Greensboro, NC 27402-6171, USA
Title: Extreme Egocentric Network Disruptions: Information Implications and Affective Outcomes
Certain events outside the control of individuals result in extreme disruptions of one’s egocentric network. Natura
disasters such as hurricanes, tsunamis, fires, and the like often result in extreme network disruptions where
contact with friends, families and others are abruptly discontinued. The result of these swift and surprising events
often leave individuals feeling isolated due to a lack of connectivity but the sense of loss goes beyond the mere
sense of isolation.

The affective outcome of receiving important information from caring providers (family, friends, neighbors, others)
can be a very powerful connector and one that is still rather under-examined in academe but not one that is
misunderstood the real world. In a previous study, Hersberger, Pettigrew and James (Social Capital as
Embedded in Social Support Networks of Homeless Populations, Sunbelt XX) found that for many of the
homeless people interviewed for the study that “forms of social capital are embedded in relations with social
Abstract: service staff, and focus primarily on access to information, tangible resources, and emotional support (italics
added) that can improve current conditions of living.” Providing needed information in a caring manner is highly
valued by information seekers in stressful situations, so much so in the homeless study that when mapping out
their social networks informants would place caring staff members in the “friends” area of the map while the
named information providers stated that they were simply “doing their jobs. “

In this project five case studies of abused and neglected children (n= 11) compare their networks prior to being
removed from their homes and placed in foster care with relatives or strangers to their networks afterward. Those
placed with relatives experienced less stress and information seeking was fairly straightforward. For those
children placed in foster care with strangers the process of information seeking was much more stressful as was
the dislocation from friends, families and neighbors. Over time, some children were able to build new support
networks but these were still tentative and sparse due to uncertainty of the outcomes of their individual cases.
Session: egocentric network
Keywords: information seeking behaviour
Accepted: Yes

183 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Julie M. Hite
Brigham Young University, Dept. of Educational Leadership & Foundations, 306J MCKB, Provo, UT 84602, USA
Authors: Steven J. Hite
Brigham Young University, Dept. of Educational Leadership & Foundations, 306E MCKB, Provo, UT 84602, USA
Biao Chang
University Preparatory School 2200 Eureka Way Redding, CA, 96001
Normative Constraints and Network Structure: Micro and Macro Co-Authoring Network Structures in the Field of
Comparative and International Education
In strategic networks of human actors, questions of agency in the creation and evolution of network structures is
a reoccurring theme (Mizruchi, 2004) which generally addresses questions of agency (proactive choice) versus
path dependence (exogenous and/or historical constraints) (Stinchcombe, 1965). Given network action may best
be modeled as a combination of choice and constraint, the theoretical problem is to delineate types of constraint
and how they interact with actor choice.
One exogenous constraint on actor agency is the presence of institutional norms within the actor’s external
environment (DiMaggio & Powell, 1983) which may make certain network choices more feasible or attractive than
others. In the context of peer-reviewed publications, micro-level coauthoring choices (network ties) made by
academics (actors) interact with macro-level publication choices (exogenous constraint) made by the institutional
field, as represented by editorial boards of the field’s academic peer-reviewed journals. The interaction of these
micro- and macro-level publication choices may create structural patterns in the co-authoring networks that
reflect institutional norms (Joyce & Lopipari, 2005). As academics increase their awareness of patterns within
co-authoring network systems, they may increase their ability to navigate strategic co-authoring network choices
in the face of existing exogenous constraints such as institutional norms.
The research question for this study is: To what extent do structural patterns exist in a field’s co-authoring
network that suggest institutional norms regarding publication? This study examined co-authorship publication
Abstract: patterns (micro) and the aggregated organizational co-authoring patterns of their affiliated institutions (macro)
within the field of Comparative and International Education. Data were collected from five of the field’s top
peer-reviewed journals published between 1994 and 2006. The analysis specifically examined micro and
macro-level factors that may suggest institutional norms. Author-level factors included gender, affiliated
institution, country, publication strategy, and publication history. Affiliated-institution factors included country,
continental clustering and ranking in the field.
Findings suggested several normative patterns in the field, including: 1) scholars tend to publish independently;
2) the field is dominated by scholars and institutions in the U.K and the U.S; and 3) these journals tend to publish
more authors from the hosting countries of the journal. No patterns were indicated regarding author gender,
publication strategy or publication history. No patterns emerged (e.g. centrality) related to the institution’s ranking
in the field.
Implications of institutional publishing norms are discussed in relation to actor choice (agency), proposing that
normative constraints may be as equally challenging as structural, relational and historical (path dependence)
constraints on an actor’s network development and evolution—in this case, the ability to successfully co-author
and publish. The interaction between micro and macro-level network choices, institutional norms, and the role of
isomorphism (DiMaggio & Powell, 1983) on the structure of the institutional network (Abrahamson & Fombrun,
1992) are also addressed.
Session: Academic Scientific Networks
co-auhorship network , structure of collaboration , network structure , network strategy , network constraints ,
international networks , micro/macro networks , actor agency
Accepted: Yes

184 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Valentina Hlebec
Authors: University of Ljubljana, Faculty of Social Sciences, Kardeljeva Pl. 5, 1000 Ljubljana, Slovenia
Gasper Koren
University of Ljubljana, Faculty of Social Sciences, Kardeljeva Pl. 5, 1000 Ljubljana, Slovenia
Title: Antonucci's Hierarchical Approach for Measuring Social Networks: self-administered application on the Web
In Antonucci's approach, emotional criteria are used for selecting alters from the respondent's global network and
placing them into three hierarchical circles that are graphically presented to the respondent. The respondent
(ego) is at the center of the three circles. The more central the circle, the closer and more important are the
people (alters) within it. The technique begins by asking the individual to look at the diagram of three concentric
circles, with a smaller circle in the center containing the word 'You'. Respondents are asked to fill the circles and
name as many alters as they want along the three criteria of importance.
Ordinarily, this approach is used with paper and pencil data collection technique. Application of this name
generator to Web self-administered questionnaire poses several questions, which we addressed using qualitative
evaluation techniques such as eye-tracking, observation and audio recording, focus groups and cognitive
interviews in two-step procedure. Findings of qualitative testing and further improvements of the questionnaire
are presented. Qualitative methods for testing questionnaires are discussed with regard to their possibility to
detect systematic measurement errors.
Session: Collecting Network Data
Keywords: methodology , method triangulation , egocentric networks , personal networks
Accepted: Yes

185 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Bernard Hogan
University of Toronto, Sociology, Department of Sociology, 725 Spadina Ave., Toronto, ON M5S2J4, Canada
Wojciech Gryc
University of Toronto, Centre for Urban and Community Studies Suite 400, 455 Spadina Ave. Toronto, ON
M5S2G8, Canada
Title: Scratching the Egocentric Itch: Designing Software for Personal Network Analysis
Personal network analysis has been a vital part of the field of network analysis since its inception. Moreover, this
style of network analysis has received its share of attention both in the social science literature and in the broade
public (consider the attention following McPherson, Smith-Lovin and Brashears 2006). Yet, there still remain a
series of technical hurdles to a robust programme of personal network analysis. That is to say, the tools for
performing analysis of multiple networks have fallen behind the tools for a single network, both in terms of
usability and analytic power.
This paper illustrates some of the unique features of personal network analysis that can be better reflected in
software. In short, we believe that personal network analysis is more than the simple batch processing of any and
all network statistics. We highlight
deisgn needs and illustrate how these issues can be implemented in a software program. We demonstrate
Abstract: "Egotistics", fully functional (beta) software that approaches personal network analysis from an entirely new
vantage point, and one we believe that will be intuitive and rewarding for technical and non-technical researchers
The core novelty of this program is its ability to build and view a multi-level dataset for personal networks in
real-time. In a series of tabs, one can view ego's attributes, alter's attributes, ego's adjacency matrix, a visual
network representation and a html-enabled console. By using the console or the menu, one can explore all
networks simultaneously, rather than one at a time. For example, it will calculate all degree centrality measures
for all alters in all networks, and append these data to the alter attributes.
We conclude by illustrating how our program employs several pre-existing tools in the social network analysis
field (such as JUNG and GraphML), and highlight the role of interfaces in making network analysis accessible to
the non-technical researcher.
Session: Software
Keywords: data representation , egocentric networks , software
Accepted: Yes

186 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jo Holliday
Authors: hollidayj1@cardiff.ac.uk
Cardiff Institute of Society, Health and Ethics, Cardiff University, 53 Park Place, Cardiff, Cardiff, Wales
Using social networks to diffuse anti-smoking messages and norms: the importance of opinion leaders’ social
The ASSIST intervention used peer-nominated, trained ‘peer supporters’ to diffuse a smoke-free message to
students in their school year. This school-based peer-led intervention was evaluated in almost 11,000 students
from 59 schools in England and Wales, and was found to be effective at one-year follow-up.

Outcome evaluation data shows that the nomination approach successfully identified peer supporters
representative of the cohort in terms of gender, ethnicity, self-reported smoking and intentions at age 16, and tha
the majority (87%) of Year 8 students knew at least one peer supporter. However, qualitative data raises
questions about whether the peer supporters represented different friendship groups and if they had the potential
to diffuse information across the year group.
This presentation builds on previous work and uses a sample of social network data gathered in ten schools
immediately post-intervention to explore these issues more directly.

Questionnaires completed by 1,779 of 1,860 eligible Year 8 (age 12-13) students elicited the names of 3,064
friends. The majority (1,788) of these friends were in Year 8 at the same school. Students were allocated to
discrete social clusters, and the degree to which risk-taking and risk-averse clusters were represented by peer
supporters was assessed. Locational properties of peer supporters will provide an indication of their potential to
disseminate the smoke-free message to their peers. The implications of these findings for the success of the
ASSIST intervention, and the potential of social network analysis to harness peer influence more successfully to
promote positive health choices will be discussed.
Session: social networks and health
Keywords: adolescence , diffusion , friendship networks , smoking , peer education
Accepted: Yes

187 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Betina Hollstein
Humboldt Universität Berlin, Institut f. Sozialwissenschaften, (Mikrosoziologie), Unter den Linden 6, Berlin D -
10099, Germany
Title: Collecting Network Data with the Narrative Interview
In this presentation the Narrative Interview (Schütze) is introduced as a tool for the collection of social network
data, which improves the quality of the collected data and which is suitable for a wide range of topics. As it will be
shown the Narrative Interview is suited for the exploration of social networks as well as for the investigation of
network interpretations, network practises and network dynamics. Furthermore it is illustrated, that the Narrative
Interview is useful for the investigation of very different network types, not only personal networks, but also intra-
and inter-organizational networks.
The paper starts out with an overview on the methodological foundations and the basic principles of the Narrative
Interview as described by Schütze and shows how to conduct it. Advantages and limitations of this method are
discussed. To compensate for certain weaknesses of the narrative interview with respect to network research
(e.g. bias on today’s relationships, time-consuming), a research strategy is proposed which combines the
Narrative Interview with more standardized name-generators (“between- and across-method-triangulation”,
Denzin) as well as with special sampling techniques. Finally, examples for these research designs are presented
illustrating how to use this method and its advantage for different fields of network research as well as for the
investigation of various network types.
Session: Qualitative Network Research: Data Collection
Keywords: qualitative research , interviews as data , data collection , mixed methods , sampling
Accepted: Yes

188 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Seok-Hee Hong
University of Sydney School of IT Sydney, Australia NICTA (National ICT Australia)
Xiaoyan Fu
Authors: NICTA (National ICT Australia) Sydney, Australia
Philippa Pattison
University of Melbourne School of Behavioural Science Melbourne, Australia
Garry Robins
University of Melbourne School of Behavioural Science Melbourne, Australia
Title: Visualisation and Interaction of Temporal Multi-relational Networks
Many social networks have multiple types of tie and change over time. However, it is difficult to use traditional 2D
visualisation methods to represent such complex series of networks.

In this abstract, we present a new framework which integrates visualisation and interaction methods in order to
help a user explore multi-relational and/or temporal networks.

Our new framework is based on 2.5D visualisation method. More specifically, we visualise each network in 2D
using a 2D drawing algorithm.
Then we arrange each 2D drawings in two parallel directions, horizontal and vertical in 3D.
In order to help users to explore the visualisation of a series of networks in 3D space, we have designed a series
of interaction methods, including selection, zooming, flipping and rotation.

By exploring a series of 2D drawings in 3D space combined with interaction methods, users can easily compare
the differences between networks,
and hence discover expected or unexpected differences.

Our framework has been implemented and evaluated using both empirical and simulated social network data
Session: Visualization
Keywords: visualization , multi-level networks , changes in networks , dynamics on networks
Accepted: Yes

189 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Cyril Hoschl
Authors: sadilkova.hana@qedquod.cz
QED & QUOD Hastalska 760/27 Prague 1, 110 00 Czech Republic
Title: Vizualization of team dynamics with Sociomapping® analysis
The research of Sociomaps was driven by a need to develop a tool pertinent to display relationships among
objects (such as network data) in user-friendly manner. Comparing to social network analysis, Sociomap
provides overall picture of a team rather than displaying all the ties between objects. Sociomap is an abstract
map-like picture of a group of objects, in which hierarchical clusters dividing the team into subgroups are
sketched. This helps to understand the team structure better.
Output of Sociomapping analysis (SMA) is not only a static picture, but can also be a dynamic animation.
Dynamic Sociomaps enable to monitor progress of system change (such as change in team relations) by
displaying the changes in the map as an animation similar to weather-forecast animation. SMA shows not only
the change in team behavior but also provides a short-time prediction of future progress.
Network data is displayed in Sociomap as a 3D landscape. Network proximities stand for landscape proximities,
but Sociomaps offer to insert additional information into the terrain - the height of the surface. This can be
arbitrary variable such as centrality degree or social status.
Practical demonstration of SMA applied to team dynamics will be presented with the help of the latest version of
Sociomapping software.
Session: N/A
Keywords: data representation , network dynamics , network structure , teams , visualization
Accepted: Yes

190 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Bettina Hoser
Universitaet Karlsruhe (TH), Information Services and Electronic Markets, Karlsruhe, Germany
Authors: David Dekker
Erasmus University Rotterdam, Department of Econometrics, Rotterdam, The Netherlands
Jan Schröder
Universitaet Karlsruhe (TH), Information Services and Electronic Markets, Karlsruhe, Germany
Title: Efficiency of Communication
Communication between individuals is costly as it consumes time and redirects attention. An important question
in business contexts is whether communication behavior of individuals in organizations is economically efficient.
This holds especially for communication intensive organizations like for example consulting companies.
Efficiency not merely depends on incoming and outgoing communication behavior but more so on the network
structure of communication patterns. A traditional measure to capture the structure of communication patterns is
centrality, for which many different measures have been developed. An often used measure is betweenness
centrality and based on previous research it may be assumed, that an actor with a high betweenness centrality
Abstract: can influence communication and outcomes of such communication more efficient. We propose to look at a
disaggregated measure, to show where within the network the betweenness of an actor can generate the highest
impact. This study focuses on the contribution of different dimensions of betweenness centrality to economic
efficiency. We use data on a mobile telephone communication network of 275 consultants and the number of
projects and the time to finishing these projects as economic performance indicators. As the standard analysis of
the communication network we used eigenvector centrality for the asymmetric network. We also found that there
seems to be a high correlation between the complex valued eigenvector centrality measure and the betweenness
Session: Communication Networks
Keywords: centrality , communication network , efficiency
Accepted: Yes

191 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Daniel Hruschka
Authors: dhrusch@santafe.edu
Santa Fe Institute, 1399 Hyde Park Road, Santa Fe, NM 87501, USA
Title: Limits on the Creation of Social Capital: a View from U.S. High Schools
Differential access to social capital has been implicated in disparities in health, economic status, and educational
attainment. However, little research has examined the processes by which the social ecologies underlying social
capital are assembled. This is an important question because a social ecology is itself constructed by motivated
individuals who must weigh trade-offs, deal with structural barriers, and make personal sacrifices in seeking and
cultivating personally adaptive relationships. This paper reports results from a one-year longitudinal study of U.S.
Abstract: high school students’ social networks (N~8000), describing key factors that constrain and enable individuals'
ability to cultivate social networks that provide particular social goods, in this case companionship and social
status. I also show how individuals appear to limit their goals in the face of such constraints, thus perpetuating
differences in social capital. Furthermore, I discuss the importance of studying how social support networks
emerge through individual action in understanding the commonly cited relationship between social capital and
other desirable outcomes, such as those related to health and economic status.
Session: Social Capital
Keywords: social capital , network change , adolescence , longitudinal , relationship formation
Accepted: Yes

192 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Yeu-Sheng Hsieh
Department of Agricultural Extension National Taiwan University No. 1, Section 4, Roosevelt Road, Taipei,
Hsuan-Tzu Chang
Department of Psychology National Taiwan University
Chyi-In Wu
Institute of Sociology Academia Sinica, Taiwan sss1ciw@gate.sinica.edu.tw
Title: The Relationships among Self-esteem, Delinquent involvement, and Network Delinquency among Adolescents
The purpose of this research is to explore the relationship among self-esteem, delinquent involvement, and
network delinquency among adolescents. Four competing hypotheses regarding the causal relationship among
self-esteem, delinquent involvement, and peer network delinquency, including self-enhancement hypothesis,
social influence hypothesis, social selection hypothesis, and continuity hypothesis, are formulated and examined
by utilizing the three-wave data of the panel study from Taiwan Youth Project (TYP). TYP is the first local panel
study of 2690 adolescents in the 7th grade, began year 2000 and is conducted by the Institute of Sociology,
Abstract: Academia Sinica, Taiwan. Applying nomination data in social network analysis to measure friends’ network
delinquency of adolescents allows for a more specific conceptualization of peer network delinquent. Self-esteem
is ditinguished to two dimensions, positive self-esteem and negative self-esteem, in this study. Both delinquency
and network delinquency are measured by three types of offense, including status offense, violent offense, and
drug abuses. Structural equation models with latent concepts are applied to assess the four proposed competing
hypotheses. The result shows that all social influence, social selection, and continuity hypotheses are expected
to explain the relationship among adolescent’s self-esteem, delinquency, and peer network delinquency.
Session: Adolescent Friendship Networks
delinquency , self-esteem , network delinquency , social influence , social selection , self-enhancement
theory , differential association
Accepted: Yes

193 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ray-May Hsung
National Chengchi University, Department of Sociology, Taipei, Taiwan
Authors: Ronald L. Breiger
University of Arizona, Department of Sociology, the United States breiger@arizona.edu
Yi-Jr Lin
Tunghai University, Department of Sociolgy, Taichung, Taiwan yijrlin@gmail.com
Title: The Creation of Differential Social Capital in Taiwan, China and the United States
How do people in different societies access social positions at all levels throughout the status order? We present
a comparative study of access to 22 positions based on survey data collected in Taiwan, China, and the United
States in 2004. The 22 positions range from lawyer and corporate chief executive officer down to security guard
and janitor. On the structural level we examine institutional logics, discovering patterns that link positions
accessed and the kinds of tie used to access them. On the individual level we examine the determinants of socia
capital quality (the status of the positions to which one is linked). Among our findings are these. Kin ties have a
negative effect on accessing positions in all three societies, whereas ties to ordinary friends have a positive effec
Abstract: in each of these societies. Good friends increase the number of accessed positions in Taiwan, but they seem to
decrease the number of total positions accessed in the United States. However, good friends are an effective
means of increasing the number of high-prestige positions accessed in the US. Work colleagues increase the
number of accessed positions in China and the US, but not in Taiwan. School ties increase the number of
high-prestige positions accessed in the US. Locality ties (i.e., having relations with others in the same local area)
are negatively related to the number of positions accessed in Taiwan but are not significant in China and the US.
Our study of the meanings and functioning of specific social ties thus leads us to identify cross-societal
similarities and differences in accessing varying qualities of social capital.
Session: Social Capital
Keywords: social capital
Accepted: Yes

194 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Mark Huisman
Authors: j.m.e.huisman@rug.nl
University of Groningen, DPMG, Heymans Institute, Grote Kruisstraat 2/1, Groningen 9712 TS, The Netherlands
Title: Imputation of missing data in social networks
Analysis of social network data is often hampered by non-response and missing data. Recent studies show the
negative effects of missing actors and ties on the structural properties of social networks, like centrality measures
and degree distributions. This means that ignoring the missing ties and analyzing the complete cases only, can
severly bias the results of social network analyses. To overcome the problems created by missing data, several
Abstract: treatment methods are proposed in the literature: model-based methods within the framework of exponential
random graph models, and imputation methods. In this paper we focus on the latter group of methods, and
investigate the use of imputation to handle missing network data. We investigate situations in which imputation
can be succesful in treating missing data, and examine the properties of some (simple) imputation methods and
their effect on structural properties of social networks.
Session: Missing Data
Keywords: missing data
Accepted: Yes

195 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

David Hunter
Penn State University, Statistics, 310 Thomas Bldg, University Park, PA 16802, USA
Authors: Martina Morris
University of Washington, Sociology and Statistics, Seattle, WA, USA
Mark Handcock
University of Washington, Statistics, Seattle, WA, USA
Title: Estimation of mixing matrices from egocentric bipartite network data
Exponential random graph (ERG) models are increasingly useful for modeling network data. The parameters in
an ERG model may be estimated based only on an estimate of the statistics sufficient to that model; it is not
necessary to observe the entire network. Methods for estimating the sufficient statistics will vary depending on
the network sampling scheme used. This talk considers the particular case of a heterosexual partnership network
in which the actors are stratified according to a certain categorical attribute (such as age) and in which the data
Abstract: are sampled egocentrically. Of particular interest is the estimation of the mean mixing matrix, by which we mean
the expected counts of relationship types broken down by the age categories of the male and the female actors,
for a population of a given age/sex profile (which is not necessarily the same as the profile in the sample). We
show how to apply earlier work on loglinear modeling (Morris, 1991), together with Poisson regression, to this
estimation problem. We apply these techniques to egocentric data on sexual networks in Uganda and King
County, Washington.
Session: Methods and Statistics
Keywords: bipartite networks , egocentric network , ergm (p*)
Accepted: Yes

196 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Rebeca E. Hwang
Authors: rebex@stanford.edu
Stanford University
Title: The Role of Social Networks on the Performance of Social Enterprises

In many urban areas in the developing world, the problem of delivering potable water to the poor has proven to
be intractable for public agencies, private companies, and NGOs. In some urban districts, the poor have gained
access to drinking water by organizing themselves into cooperatives, voluntary membership organizations that
are democratic and member controlled, and that have a "double bottom line": they must remain financially
sustainable and deliver water to all their members, including the poor. These two goals are often conflicting. This
paper presents preliminary results from a study that investigates propositions concerning how social networks
among members of cooperatives, and affiliations between leaders of cooperatives and local elites, affect the
ability of cooperatives to meet their dual goals in Argentina. This research draws theoretical insights from social
network theory and from literature on water service delivery and social entrepreneurship in developing countries.

This study is located in Argentina. Therefore, a thorough description of the field of water service delivery and
water cooperatives in Argentina will be presented for this ongoing study. Hundreds of small-scale water
cooperatives coexist with other types of service providers in the the main metropolitan areas. This presentation
will provide a typology of water cooperatives based on, among other factors, on their internal regulations,
decision-making processes, internal network architecture, external affiliation patterns and aspects of their social
capital. In addition, highlights for the interdisciplinary research design used in this study will be included,
particularly those focused on the metrics used to measure performance given different network structures.
Finally, the results for the first case study will be presented, as well as the hypotheses and future research plans
supported by this case study.

This interdisciplinary project contributes to ongoing discussions in the literature regarding: the role of small scale
water providers in improving access of the poor to water; the influence of social capital on the delivery of water
services in growing, informal urban settlements; the effect of specific social network structures on the
achievement of goals of enterprises that have both a financial and a social mission; and the creation of a
performance measure of water service providers that include a component on access of the poor to water.
Collectively, the results can inform policies intended to enhance the availability of water services for the urban
poor, and provide insights on the effect of specific social network architectures in the performance of an
Session: Performance Metrics for Social Networks
functioning of social capital , improving business performance , interdisciplinary approach , effectiveness ,
water cooperatives
Accepted: Yes

197 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sungsoo Hwang
University of Pittsburgh, Graduate School of Public and International Affairs, 121 University Pl Room301,
Pittsburgh, PA 15260, USA
Past, Present, and the Future of Social Network Analysis: Network as a Metaphor, Paradigm, Theory, or
In brief, this paper has two basic objectives. First, it intends to address how SNA is perceived by various
disciplines and what that means in an academic world, as being rooted in a hermeneutics tradition. Second, it
aims to forecast future directions of SNA and network.

This work is inspired by Steve Borgatti’s session, ‘What is network theory?’ at the Sunbelt 2006. He talked about
many social science theories and research traditions that came from network perspectives, and there was a
subtle implication that it has not been noticed as such yet. “Network (or Social Network)” is everywhere today, ye
its presence in various academic disciplines differs. In January 2006, there was much debate whether network
analysis is a theory or not in ‘socnet’ as well.

This paper will briefly track the history of SNA and discuss this. Qualitative techniques are utilized to identify the
issues. Using computer assisted text analysis tools such as Atlas.ti and Automap, this paper qualitatively codes
and analyzes the discussion on the socnet and from the author’s survey.
The second focal point of this paper is to have judgmental forecasts using Delphi technique. The survey is being
conducted to get ‘expert opinions’ on the future of the SNA. The paper aims to forecast where SNA would go in
the next decade and the next generation.
Session: Ontology and Philosophy of Networks
Keywords: network theory , forecasting , delphi technique , social network
Accepted: Yes

198 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Tasuku Igarashi
Osaka University Department of Social Psychology Osaka, Japan
Authors: Garry Robins
University of Melbourne Department of Psychology Melbourne, Australia
Pip Pattison
University of Melbourne Department of Psychology Melbourne, Australia
Title: Dynamics of social networks, trust, and social identity
Social networks change over time. The exponential random graph model was applied to longitudinal network data
in order to capture the dynamics of network structure, trust and social identity. The data consisted of 70 first year
undergraduates at an education department of a university in Japan. They nominated friends at the department
whom they had greeted and/or with whom they had discussed personal matters face-to-face and/or via mobile
phone text messages at four time points (T1 to T4) during their school year. Generalized trust toward others,
identification with the department, and identification with the department friends were assessed as psychological
dispositions. Undergraduates with high generalized trust were less likely to nominate friends throughout the first
Abstract: semester (from T1 to T2), but this trend was reversed throughout the summer vacation (from T2 to T3): those
with high generalized trust were more likely to nominate others than those with low trust. These results imply the
role of trust as social intelligence: people with high generalized trust might be active in relationship formation, but
they might be careful to form relationships at an initial stage so as to distinguish others to be trustworthy or not.
In the vacation period, when friendships seemed to be reassessed, increasing departmental identification was
associated with more friendship nominations. In the same period, increasing identification with friends was
accompanied by more face-to-face nominations, but fewer text message nominations. The implication of
'relationship-reassessment' during the vacation will be discussed.
Session: Exponential Random Graphs
Keywords: ergm (p*) , friendship networks , dynamics on networks , trust
Accepted: Yes

199 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Olivia Ip
City University of Hong Kong, Management, Department of Management, City University of Hong, Kong, 88, Tat
Authors: Chee Avenue, Kowloon, Hong Kong SAR, China
Sek Hong Ng
Faculty of Business University of Hong Kong Pok Fu Lam Rd., HKSAR.
Social connectedness of occupational community and work behavior: a study of the social relations of
construction workers in Hong Kong
Traditional occupational communities in the past have always been sustained on a ‘dense’/close social network
among members belonging to the same trade or occupation. The homogeneity and cohesiveness among
members facilitate transmission of job information and knowledge, and exert effective control over the standard
of work within a trade or occupation. Examining social relations in an occupational community allows for study of
work relations outside of the ‘bureaucratically’ and ‘vertically organized’ work organization. Other than this, a core
feature of occupational community is the pervasive embrace of the occupational identity of the individual: the
occupational identity serves a key source of identity and embeddedness for its members, leading to higher
motivation to exert effort across organizations. Hence, it is proposed that there is strong relationship between
social connectedness of members of work groups in occupational community and work behavior. So far,
research is sparse in this area.

The study sets out to fill in the gap and attempts to explore the following questions:
1. What are the relationships between patterns of social relations among members of an occupational
community, and occupational identity?
2. What are the relationships between patterns of social relations among members of an occupational community
and work behavior?

For the above purpose, and in view of the exploratory and complex nature of the study, we employ a qualitative
approach. Work groups in a construction site are interviewed to collect data on the research questions. The
construction industry is selected because of the comparatively low degree of ‘vertical’ organization in the
industry, which is dominated by ‘out-sourcing’ practice. Hence, most of the workers are not directly employed by
the construction firm, but only work on a project basis. Through employing a qualitative approach, we have
emphasized the contextual factors within which the networks of the construction workers are embedded. More
details of the findings can be presented upon completion of analysis.
Session: Qualitative Network Studies I
Keywords: occupational community , social identity
Accepted: Yes

200 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Zachary Jacobson
School of Computer Science, Carleton University, Ottawa, Canada and Health Policy Research Division, Health
Authors: Canada
Dragos Calitoiu
Health Policy Research Division, Health Canada, Ottawa, Canada
Title: A co-authorship network: the analysis of the structure of scientific collaboration
Scientific co-authorship implies a temporal and collegial relationship; co-authorship networks have been studied
to infer the social structure of the scientific collaborations and statuses of individual researchers. The famous
contribution to this field is the Erdos Number Project in which the number of co-authorship links between any
mathematician and the Hungarian mathematician Erdos was calculated. Co-authorship analysis has been
applied to numerous ACM Conferences, e.g. Information Retrieval (SIGR)and Management of Data (SIGMOD).

We will report authorship collaboration among professors in the School of Computer Science, Carleton University
(Ottawa, Canada). In order to generate a co-authorship network, we counted all the papers registered in the
Digital Bibliography & Library Project (DBLP) database as of July 2006 written by the faculty of the School of
Computer Science. The DBLP database contains scientific journals and proceedings from the field of Computer
Science and it is administered by Computer Science Bibliography University Trier (dblp.uni-trier.de). It lists more
than 830000 articles. Students (undergraduate or graduate) or Post-Doctoral Fellows were not included in our

The results obtained describe the structure of the collaboration in this research community and quantified the
status and the influence on the actors in the network. Our future work will be to compare the structure of scientific
collaboration across different disciplines at our (Carleton) University with corresponding structure of collaboration
in the same discipline but different universities.
Session: Academic Scientific Networks
Keywords: co-auhorship network , structure of collaboration
Accepted: Yes

201 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Lorien Jasny
Authors: ljasny@uci.edu
UC Irvine, Sociology, 4101 Palo Verde Rd, Irvine, California 92617, USA
Title: Networks of Belief Structures: An Entailment Analysis of the ISSP 2003 National Identity Survey
This paper presents an examination of the ISSP 2003 National Identity Survey using the network entailment
analysis. The questions in the survey are similar to those in many other multinational surveys. Examples include
questions about the individual’s pride in different aspects of the country, attachment to institutions, and
international relations. The entailment process uses thresholds of logical implication to extract cognitive
structures in the relationships between the questions asked. For example, when a majority of respondents who
answer positively to question A also answer positively to question B, such an entailment relationship is observed.
The pooled results, of 48 different questions, form signed networks of dyadic relations between the 48 nodes
indicating patterns of positive and negative responses. These cognitive network graphs are compared over each
of the 33 countries included in the ISSP study, and the countries are clustered by similarities in the entailment
networks extracted. Finally, the entailment structure permits an examination of the common relationships among
responses within the clusters through extracting the central graphs. This process compares not only the
dominant response patterns within countries, but between countries as well. Thus, this method produces a new
way of looking at multinational survey data that incorporates micro level differences in cognitive structure into a
macro level comparison.
Session: Cognitive associative networks
Keywords: political culture , cognitive networks
Accepted: Yes

202 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Kent Wickstrøm Jensen

University of Southern Denmark, Entrepreneurship and Relationship Management, Engstien 1, Kolding, Jylland
6000, Denmark
Title: Underperforming but unaware: a matter of position?
Underperforming but unaware: a matter of position?
Does network position affect the gap between subordinate self-assessment and supervisor evaluation of
subordinate task performance? Network position may affect this gap in several ways. First, network position
affects the subordinate’s access to information about the criteria by which performance is evaluated, how well the
subordinate performs relative to these criteria, and how well the subordinate performs relative to co-workers.
Secondly, the network position of the subordinate may work to form the supervisor’s expectation about what can
be expected from the subordinate. It can by hypothesized that subordinates occupying central positions will be
well informed but also have to live up to higher expectations. In a study of 4 product-development projects in a
major Danish IT company, task performance of project task were evaluated by both the project participants
allocated to the tasks and the associated project manager. The results indicate that the propensity of
subordinates to overestimate performance increases as the geodesic distance between the subordinate and the
supervisor increases, and as the popularity of the subordinate in the advice network increases. Seemingly,
neither a “out of sight” nor a “here I am” strategy pays of well when it comes to exceeding supervisor
Session: Intra-Organizational Networks
Keywords: individual performance , advice , egocentric networks
Accepted: Yes

203 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Catherine A. Johnson
University of Wisconsin-Milwaukee, School of Information Studies, 586 Bolton Hall, 3210 N. Maryland Avenue,
Milwaukee, Wisconsin 53211, USA
Title: The public library: A repository of social capital?
This presentation reports on a study that investigated the relationship between use of public libraries and social
capital. Two concepts of social capital, as a collective resource (Putnam, 1995) and as an individual resource
(Lin, 2001), frame the study. The public library is an important social institution since it is a public space where
people of the community gather, meet one another and freely access community resources. Putnam has implied
that social capital emerges from social interactions of the kind taking place in libraries, suggesting that libraries
may be an important resource for helping to build community cohesion. Lin’s concept of social capital as an
Abstract: individual resource, however, suggests that libraries may also be an alternative for people with low levels of
social capital since it is a place where they can make use of information technology and the knowledge and help
of librarians when they have no one in their personal networks who can provide the resources they need.
Questionnaires were administered to 120 public library users and interviews were conducted with 15 library staff
members at three branch libraries in poor neighborhoods in a large Midwestern American city. Findings from the
study will not only show the relationship between libraries and social capital, but will also indicate whether the
relationship is similar for both concepts of social capital.
Session: Social Capital
Keywords: access to knowledge , methodology , social capital , position generator , public space
Accepted: Yes

204 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jeffrey Johnson
East Carolina University, Sociology, ICMR, Greenville, NC 27858, USA
Terry Reynolds
Authors: ICMR East Carolina University Greenville, NC 27858, USA
David Griffith
ICMR East Carolina University Greenville, NC 27858, USA
Lisa Clough
Delartment of Biology East Carolina University Greenville, NC 27858, USA
Ecological Knowledge, Knowledge Reputation and Social Networks among the Iñupiaq of the Northern Kotzebue
This paper examines the relationship between various forms of ecological knowledge, often referred to as
traditional ecological knowledge, (TEK), knowledge reputation, and aspects of social networks among Iñupiaq
marine mammal hunters of Northwest arctic Alaska. The study consists of interviews with 81 Inupiaq subsistence
hunters considered the most active based on subsistence surveys by the tribal association. Data on who are
knowledgeable ugruk (bearded seal) hunters, who hunters talk to about marine mammal hunting, knowledge of
Abstract: the Northern Kotzebue Sound ecosystem and marine mammal behavior, knowledge of Kotzebue Sound trophic
networks, and hunter characteristics were collected. The study finds a clear and strong relationship between
knowledge reputation and various aspects of the hunter’s communication network, particularly indegree centrality
and network core membership. However, hunter’s ecological knowledge itself is a more complex set of
relationships among not only network factors, but also factors such as elder status, formal education,
dependency on subsistence products, and village residence.
Session: Knowledge Networks
Keywords: knowledge networks , intracultural variation , cultural ecological models , perception
Accepted: Yes

205 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Lisa Johnston
Authors: lsjohnston@comcast.net
Independent consultant, Tulane Univeristy, School of Public Health and Tropical Medicine
Formative Research for Using Respondent Driven Sampling in HIV Biologic/Behavioral Studies: Understanding
your Target Population
Researchers, policy makers and governmental agencies around the world are encouraged by the recent reports
of the successful use of Respondent Driven Sampling (RDS) to study HIV prevalence and risk behaviors among
vulnerable and hard to reach populations. RDS uses principles of social network theories to mitigate the biases
found in popular and easy to use chain referral samples. Unlike traditional random sampling methods, RDS does
not require a sampling frame to calculate probability of selecting a sampling unit. RDS has been particularly
useful for studying HIV vulnerable populations who practice illegal and/or stigmatized behaviors and often lack
sampling frames such as sex workers (SWs), injecting drug users (IDUs) and males who have sex with males
(MSM). It is said that one advantage of RDS is that little formative research is required. However, researchers
who conduct RDS in international settings find that formative research is an essential step in ensuring a
successful outcome for the study.
Based on the first author’s past experience with RDS, the proper formative research could have avoided
problems associated with 1) missing important members of the target population (e.g., older “sugar daddies” in
South Africa, female IDUs in Vietnam, deaf SWs in India); 2) insufficient social networking among members of
the target population (e.g., SWs in Montenegro, MSM/SWs in Eastern Caribbean) and 3) study logistics (e.g.,
incentive level in Serbia, interview site location in Croatia).

Formative research could likely improve the efficiency and success of RDS in an untested population by
identifying potential sampling issues (e.g., population’s average network size, network density), resolving study
logistics (e.g., incentive level, appropriate interview site locations) and better understanding the population (types
of sex workers, appropriate types of testing or clinical services, acceptability of an RDS study).
Session: Qualitative Network Research: Data Collection
Keywords: social networks , peer networks , qualitative research
Accepted: Yes

206 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Lukasz Jonak
Authors: lukasz@jonak.info
Warsaw University, Institute of Sociology, Karowa 18, Warsaw 00-324, Poland
Title: Between efficiency and innovativeness: the dynamics of an organization as a social system
A business organization can be conceptualized as a social system, with internal network of relations between
employees, or other units, defining the structure of the system. Firm’s operations depend in turn on the nature of
the business, but also on the relations of the company with environment – clients and competition. The internal
structure of a firm should reflect both the type of organization’s business and the structure of its market

This paper presents the evolution of internal network of IT consulting company leading to the significant
Abstract: improvement of its performance. The longitudinal network study of employees project affiliations and direct
communication networks was performed. The changes of cohesion and homogeneity/differentiation indexes were
measured in order to assess the direction of organizational change. The dimensions of this assessment are
efficiency (responsible for general business performance) and innovativeness (defining the adaptive potential of
the company, given the rapid changes in IT business trends and evolution of market structure).

The results are interpreted in the context of organizational network analysis, the structural holes–closure
argument, and generalized on the level of (social) systems approach.
Session: Intra-Organizational Networks
Keywords: innovation , intraorganizational networks , social systems , network dynamics
Accepted: Yes

207 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Yuval Kalish
Authors: Department of Sociology and Anthropology, Hebrew University, Jerusalem, Israel
Garry Robins
Department of Psychology University of Melbourne Parkville 3010 Australia
Title: Social settings and the organization of network ties: an empirical examination of Foci of Activity theory
There has been a renewed theoretical interest in the way individual differences and contextual effects are related
to network structure. The current paper investigates how contextual effects enable or constrain the development
of simple network structures. We utilize Foci of Activity Theory (Feld, 1981) and provide an empirical examination
of the specific predictions of the theory. We place Foci of Activity theory in a larger framework of theoretical
accounts that pertain to the self-organization of network ties into transitive triads. We propose extensions to the
definition of foci characteristics, including psychological intensity, importance, compatibility and the presence of
Abstract: conflict. We argue that there are two corollaries to the theory relating to, firstly, the development of network ties
within foci of activity and, secondly, to the development of ties between foci and the emergence of new foci from
old ones. In examining these two aspects in separate analyses, we find variability in the tendency of transitivity
based on the characteristics of foci. Moreover, our results also supported Feld’s mechanism of focus integration:
ties were more likely to occur between social settings that were viewed as compatible. We close with a
discussion of the findings, including how these results are consistent with new proposals for dependence
hypotheses for statistical models for social networks.
Session: Network Dynamics
Keywords: affiliation networks , ergm (p*) , transitivity , personal networks , network structure
Accepted: Yes

208 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Stella Kalogeropoulou
Title: Social Care Networks:new forms of Social Policy in the Greek legal context
This paper examines the role of the civic social state as a Social Care Network. The focal point is the
conceptualisation of civic social state as a social -and legal system of a 'becoming identity' according to the
Abstract: systemic theoretical approach of Luhmann. In this form, the social state is shaped by, and in the same time
shapes, social reality and-through this structural dynamic approach -contributes decisively to the social
Session: Social Networking Tools
qualitative network , collective action , dynamics on networks , identity , inequality , network structure , social
Accepted: Undetermined

209 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Annelies Kamran
Authors: akamran@gc.cuny.edu
Graduate Center/CUNY, Political Science, 2769 Montauk Hwy., Brookhaven, New York 11719, USA
Title: Structure of a Transnational Network: The Response to Nuclear Proliferation
The aim of this paper is to use the methods of social network analysis to discover the structural patterns of
cooperation that arise in response to a global security problem. Social network analysis methods allow one to
quantify relationships between actors and represent them mathematically and graphically, thus gaining new
insights into the intersubjective structures created by actors. By mapping the relationships at the system level of
world politics including individual states as well as intergovernmental organizations, transnational
Abstract: nongovernmental organizations and transnational businesses, this paper will demonstrate how a new way of
thinking about the constitution of system level world politics can produce new information. The theory used for
the investigation is: due to possibilities raised by globalization, the situational framework in large part determines
the network that is formed in response to a global security problem. This paper will therefore conduct a social
network analysis mapping the construction of a traditional security transnational cooperative response network,
using as a case study the response to nuclear proliferation since the end of the Cold War.
Session: International Networks
globalization , governance , inter-organizational , international networks , network structure , world systems ,
security networks
Accepted: Yes

210 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Esmail KaramiDehkordi
PhD Researcher in School of Agriculture, Policy and Developmet, The University of Reading, Reading RG6 6AR,
Authors: United Kingdom Lecturer in The University of Zanjan, Iran
Chris Garforth
Professor of Agricultural Extension and Rural Development, The University of Reading, UK
Communication and Information Networks of Social Actors in Natural Resource Management: A Case Study in
Chaharmahal and Bakhtiari Province in Iran
Strong relationship, information exchange and collaboration of social actors are essential for sustainable NRM
and rural livelihoods systems. The purpose is to investigate information exchange networks of social actors of
different institutions relevant to NRM. It is also to discuss the influence or leadership of these institutions for
collaborative actions. A case study using quantitative and qualitative data collection and analysis techniques was
utilised in relevant institutions of Chaharmahal and Bakhtiari Province in Iran.
The networks showed a small number of strong relations, but a great number of weak or no relations, which
caused a low density of the network with strong values. So information exchange of many actors was in an
irregular and infrequent way.
Abstract: Centrality measures revealed not much power and influence in information network, though centrality of the
actors was not equally distributed. Institutions’ managers and provincial and district level extension service staff
had relatively a more central and powerful position in the network. This shows a potential to manage the
complexities of the network in the future. In the same way, poor centrality and information exchange (especially
among researchers and many subject matter specialists outside extension service) might make some obstacles.
Clique, cluster, factor analysis and qualitative data identified the association of information exchange and factors
such as nature and policies of institutions; and administrative, geographical and managerial position of actors.
The strongest relations were among actors of each institution in a geographical level, while the least frequent and
weakest communication was between governmental and non-governmental actors.
Session: Knowledge and Information Networks
natural resource management , knowledge networks , information networks , knowledge systems , information
exchange , iran , collective action , rural communities
Accepted: Yes

211 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ioannis Karidas
Authors: jkaridas@altecnet.gr
Panteion University, Social Psycology, Athens Greece.
Title: New neighborhoold communities, a space-social topology of the global/ virual cities
In this paper, I describe neighborhoold communities as social relations that occure in three different levels of
space or different topologies, region, global netwoorks, virtual fluides. I try to develop a way of describing the
relations between these different space/social environments. So the critical point, the basic aim, is to find a way
of describing a method of investigate the social systems, also the relations of different forms of spatiality and so
between different versions of social relations, specially of neighborhoold/localities - cities/globalities, of social
connections in the compex envirtonments of global/ virtual cities. In this paper i suggest that these topologies of
the different spases is genererative not only of the local communities but also of the global differences between
topologies themselvs. In this way i try to develop a tool for understanding the relations between the local and the
global and the real and the virtual. This approach is strengthened by the new technologies of information and
communication and, also, by the Internet and virtual communities, which affected considerably the new social
environment. Finally, in the conclusion of these paper i describe these space /social topology from local/
neighborhoold communities to virtual/ global cities as the topology of digital cities. The main question that this
paper attemps to answer is whether similar topologies can be applied today in the existing framework of social
Session: Neighborhood Communities
Keywords: social networks , neighborhood communities
Accepted: Undetermined

212 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Harald Katzmair
FAS.research 1168 Folsom Street #304 94103 San Francsico, USA
Ruth Pfosser
Authors: FAS.research Muellnergasse 3/1 1090 Wien, Austria
Christian Gulas
FAS.research Muellnergasse 3/1 1090 Wien, Austria
Max Ruhri
Evaluating networks: Assessing the Robustness, Efficiency and Diversity of Networks by the Example of
"Interregional Cooperation" in the European Union
Dozens of highly standardized social science based methods and tools are
available for program evaluation purposes, but currently none of those
standardized approaches actually use SNA based measures. Our challenge was
to develop a plausible concept and framework to utilize SNA as
an evaluation method in the context of EU program evaluation and as a
basis for decision-making in the future design of EU-programs and

Like in every evaluation process one has to define an objective function

to determine the quality and value of the outcome. But what does it mean
to define a value function in the context of network structures? And
what if the outcome isn't a measurable standard variable (regional
Abstract: GDP, number of new start ups, number of created jobs etc.) but the
network structure (it's robustness, efficiency and diversity) itself?
What are "good" and what are "bad" network patterns? A lot of studies in
the field of organizational networks have shown that there isn't
anything out there like THE excellent network morphology. Form follows
function. Ideal network structures differ according to their goals and
tasks. Only in respect to specific goals one can speak about "ideal" or
"excellent" networks.

In this paper we will present our attempt in solving the puzzle of the
form-follows-function problem in the context of network evaluation.
Finally we want to discuss some more general scholarly and practical
questions and ambiguities which arise in using SNA as an evaluation tool.
Session: Network Evaluation
affiliation networks , eu policy making , knowledge transfer , methods , network comparison , performance
measurement , policy network
Accepted: Yes

213 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sylvia Keim
Max-Planck-Institute for Demographic Research Independent Research Group on the Culture of Reproduction
Rostock, Germany
Title: The Influence of Social Networks on the Timing of Family Formation
This paper presents the results of a mixed-methods study on the influence of social networks on the decision to
have a first or second child. We conducted qualitative interviews with young adults that centered on the topics of
family formation and social relations. Additionally the respondents’ social networks were mapped with a
network-chart and grid, which provides rich data on the individual relations as well as systematic information on
the structure of the network. Whereas most studies on ego-centered networks are based on information by Ego
only, we also interviewed up to three network partners of our main respondents in order to gain deeper insights
into interpersonal influences based on the information from both network-partners.
In this paper we focus on the respondents’ ideas on timing of family formation or extension and how this is
negotiated and transmitted in their social networks. The main questions are:
• What ideas on “timing” do the respondents present? Do they refer to their network-partners when stating these
• How do the people in the respondents’ social network deal with the timing of family formation or extension? How
do these network-partners influence the respondents’ views? What is the role of the network structure for this
Session: Qualitative Network Studies - II
Keywords: family , fertility , friendship networks , qualitative network
Accepted: Yes

214 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Natasa Kejzar
Authors: University of Ljubljana, Faculty of Social Sciences, Kardeljeva ploscad 5, Ljubljana, Slovenia 1000, Slovenia
Vladimir Batagelj
University of Ljubljana Faculty of Mathematics and Physics, Jadranska 19, 1000 Ljubljana, Slovenia
Title: Probalistic inductive classes of graphs
The idea of mathematical induction is well known for some centuries. It can be used to prove for a statement to
hold or to define a certain class of objects. To define an inductive class of objects we have to give (1) a class of
initial objects, and (2) a list of generating rules that transform object(s) already in the class into an object also in
the class. The inductive class consists exactly of the objects that can be obtained from the initial objects in finite
number of steps using the generating rules. Eberhard was the first one to define classes of graphs using an
inductive definition. In graph theory the inductive definitions for several classes of graphs were given.

Abstract: In our work we think of a graph as a ``skeleton" of the network. With inductive definitions of graphs one can
describe the evolution of a graph in some prescribed manners. The transitions (transformations by rules) can be
viewed as implicit time steps. We extend the standard notion of inductive classes of graphs (ICGs) by imposing a
probability space in the choice of rules and their initial elements and call it probabilistic ICGs (PICGs).

Many of the existing models of growing networks can be cast as PICGs.We will look at some graph/network
generations according to specific PICG rules which serve as models for specific real-world networks (i.e. spread
of knowledge and inovation among people, changing the structure of links in a fixed social network in time etc.).
Session: Mathematical Models
Keywords: network evolution , network dynamics , methodology , longitudinal , simulation
Accepted: Yes

215 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Patrick Kenis
Authors: p.kenis@uvt.nl
Department of Organisation Studies Tilburg University the Netherlands
Title: The use of network analysis in studying the production system of Open Innovation Communities
In the present paper we develop a conceptual framework for assessing open innovation communities. Open
innovation communities are a form of user-led innovation. They involve loosely regulated collaboration of differen
actors, which leads to the creation or modification of both tangible and intangible goods and services. The
defining characteristic of open innovation is collaboration and sharing in a dispersed way through the use of ICT
networks. Although open innovation is most often associated with open source software, open innovation as a
form of production is being extended more and more areas (e.g. information provision, drugs development, risk
reduction applications, the Tropical Disease Initiative, etc.).
We are interested in assessing the quality of goods produced in open innovation communities using network
analysis. We suggest that the social structure of the production system is a good proxy for the quality of the
goods produced and propose a research design for testing the conceptual framework presented.
Session: Innovation
Keywords: open innovation , network structure , effectiveness
Accepted: Yes

216 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jang Hyun Kim

Authors: SUNY@Buffalo, Communication, Department of Communication, SUNY @Buffalo, Buffalo, NY 14260, USA
George Barnett
SUNY@Buffalo, Communication, Department of Communication, SUNY @Buffalo, Buffalo, NY 14260, USA
Title: Examining the Structure and Influence of the U.S. Senate’s Hyperlink Network on Roll Call Voting Patterns
Hyperlinks among the web sites of the members of the 109th United States Senate were examined through
social network analysis. Structurally, the hyperlink network’s most central actors were primarily Democrats and
moderate Republicans. Party affiliation was a significant predictor of the hyperlink network’s structure (r = .038, p
< .005). As expected, the roll call voting network was composed of two cliques, one for each political party with
moderates from both parties bridging the two groups. While political party (r = .877, p < .005) accounted for a
large proportion of the variance in voting patterns, the hyperlink network (r = .070, p < .005) independently
explained additional variance. Together, these two variables accounted over 35% of the variance in voting
patterns. Suggestions for future research are addressed.
Session: Political Networks
Keywords: collective action , political networks , internet/hyperlink networks
Accepted: Yes

217 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Don Kincaid
Authors: lkincaid@jhuccp.org
Johns Hopkins University, Communication, 10241 Tuscany Road, Ellicott City, Maryland 21042, USA
Title: Bounded Normative Influence: The Strength of Strong Ties in Social Networks
Weak ties in social networks provide bridges over which innovations can cross the boundaries of social
subgroups, but once this occurs then strong ties within these subgroups can inhibit change by social sanctions
and reinforcement of existing social norms. This implies that weak ties favor—are necessary for—innovation
while strong ties only serve to inhibit innovation within social networks. So, how can any innovation ever diffuse to
the point where it becomes a new social norm? The paradox of how a minority can ever overcome the influence
of the majority cannot be overcome easily without taking into account the role of boundaries within social
networks. Computer simulations of social networks have demonstrated a new principle of social change that
Abstract: resolves this paradox. A new minority can change the prevailing social norm by means bounded normative
influence: the tendency of social norms to influence behavior within relatively bounded, local subgroups of a
social system rather than the system as a whole.
Clique boundaries in networks function to protect “deviant” behavior of a minority by insulating it from majority
influence. As long as a minority maintains its majority status within its own, locally bounded portion of the
network, then it can survive and even grow by recruiting converts in the near surround by means of weak ties.
This principle is illustrated by computer simulations, mothers clubs in Korea, the eradication of foot binding in
China and female circumcision in Africa, and the spiral of silence theory of political campaigns in Germany.
Session: Network Theory
clustering , cultural convergence , network change , social influence , model-based simulation , entropy-like
measures , strength of strong ties , bounded normative influence
Accepted: Yes

218 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Olga Kioufegi
University of Patras Department of Mathematics Patras, Greece
Moses Boudourides
Authors: University of Patras Department of Mathematics Patras, Greece
David Tindall
Department of Sociology University of British Columbia Vancouver, Canada
Manos Kokorakis
University of Patras Department of Mathematics Patras, Greece
Are structural and positional embeddedness in multi-organizational networks associated with concern about
climate change?
This paper examines survey data collected through the “Forests, Values, and Communities Survey” from the
general public in British Columbia in 2005-2006. We examine people’s membership in voluntary organizations,
classifying these organizations into a set of organization types. We create affiliation networks based on these
organizational types, and compute different network properties for each organizational type (positional and
structural embeddedness, membership in block models, etc.). We explore differences in the corresponding
affiliation networks between those who are concerned about climate change and those who are not, and those
who plan to do something about climate change and those who do not. Our objective is to examine the potential
correlations between structural and positional location in a multi-organizational network and people’s concern
about, and strategies for dealing with, climate change.
Theoretical implications of these findings are discussed.
Session: Policy Networks on Climate Change
Keywords: affiliation networks , embeddedness , social influence , environmental networks
Accepted: Yes

219 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Deirdre Kirke
National University of Ireland, Maynooth, Sociology, Maynooth,, Co. Kildare, Dublin, Ireland
Authors: Pip Pattison
Department of Psychology, University of Melbourne Australia
Garry Robins
Department of Psychology, University of Melbourne Australia
Title: Modelling the role of gender in structuring adolescent friendship networks
Although gender homophily is a ubiquitous feature of friendship networks
at all ages, the nature of the association between gender and network
macrostructure is not fully understood. In this paper, we use social
selection models (Robins, Elliott & Pattison, 2001) in the
exponential random graph modelling framework to assess the role of gender
in structuring adolescent friendship networks. These models allow us to
assess a variety of potential gender-related influences on network
properties, including: biases in expansiveness and popularity;
homophily and differential homophily effects; differential variation in
expansiveness and popularity biases; differential propensities for
clustering; and differential forms of cross-gender "brokerage". We
apply this modelling approach to a network of 1188 ties among 267
teenagers and their friends in an Irish community and discuss the
implications of the results for theories of network evolution.

Reference: Robins, G., Elliott, P., & Pattison, P. (2001). Network

models for social selection processes. Social Networks, 23, 1-30.

Keywords: ergm (p*) models, realization dependence,

statistical models, friendship networks, gender, homophily, clustering.
Session: Adolescent Friendship Networks
Keywords: ergm (p*) , realization dependence , statistical models , friendship networks , gender , homophily , clustering
Accepted: Yes

220 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Akishige Kishida
Ark Internationals, Kobe, Japan
John Boyd
University of California at Irvine, Department of Anthropology, School of Social Sciences, Irvine,, CA
92697-5100, USA
Title: A Random Graph Approach in Structural Balance
We present a random graph model that makes it possible to analyze data based on signed digraphs. We create
a balance index, using the first and second moments of a random signed digraph. The model is an extension of
earlier models, e.g., positive transitivity (Holland & Leinhardt, 1970), but has some advantages over them. It is
easy to calculate and interpret while fitting the data better. It is more general than the previous model in that it
can analyze signed digraphs. In our model the smallest unit of analysis is a signed, directed triad whereas it was
a directed triad in previous models. A balanced triad is defined as follows: If the product of the signs in every
semicycle is positive, then the triad is balanced; otherwise, it is unbalanced. In actual data, some triads may be
balanced while others are not. We calculate the mean and standard deviation of the number of balanced triads in
a random signed digraph, conditioned on the number of points and positive and negative arcs. The balance index
is the number of balanced triads standardized by the mean and standard deviation of this number for the
corresponding random signed digraph. The balance index measures the tendency toward, or away from,
structural balance. Furthermore it allows us to compare indices against each other in a meaningful way. The
balance index appears to be normally distributed as a result of our computer simulation. We compute this index
on the Sampson monastery data.
Session: Methods and Statistics
Keywords: structural balance , statistical models , signed digraph , random graph , simulation
Accepted: Yes

221 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sandra Kliknaite
Baltic Business School, Kalmar, Sweden and Lund University, Business Administration Department, Lund,
Title: Managing the Balancing Act between Exploration and Exploitation in Industry-University R&D Collaborations
The main question in the paper is: How can the different organizational requirements for exploration versus
exploitation be addressed to secure both creation and implementation of innovation? I believe that this issue can
be solved through new forms of networking between internal exploitation and external exploration. According to
Harryson (2006) innovation requires management of both weak and strong ties cutting across both peripheral
and core networks with a strong focus on developing and managing relationships for transfer and transformation
of information into innovation. The purpose of my research is to explore how to connect these two network
structures and, in particular, how to transfer results from exploration networks to exploitation networks. In our
previous research on a selection of technology innovation leadership companies in Europe (Harryson, Kliknaite
and Dudkowski, forthcoming) we find that external sourcing of technologies and skills does not have to result in a
Abstract: hollowing out of internal R&D capabilities. In contrast, it seems that certain mdels of Industry-University
collaboration create synergies that support the transformation of scientific knowledge and disruptive technologies
into innovation.
My research aims to take these finding further, by using and improving the following analytical framework three
interrelated network levels with different foci:
1. Extracorporate creativity networks
2. Intracorporate process networks
3. Project networks
I see a strong need to understand the role of project networks to analyze when and how which types of ties and
relationships contribute, respectively, to exploration and exploitation of technology innovation.
Session: Innovation
balance , access to knowledge , collaboration , innovation , inter-organizational networks , intra-firm
networks , tie strength
Accepted: Yes

222 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Stefan Klink
University of Karlsruhe (TH) Institute of Applied Informatics and Formal Description Methods Karlsruhe, Germany
Patrick Reuther
University of Trier Department of Databases and Information Systems (DBIS) Trier, Germany
Alexander Weber
Title: Finding Relevant Publications in Bibliographical Databases with Social Networks
Finding relevant publications to a certain topic is the classical information retrieval problem. Although good
retrieval tools exist solving this problem is getting harder and harder due to the mass of information and the rapid
growth of the number of scientific workers.
The io-port.net portal and the DBLP Computer Science Bibliography include more than 2,000,000 and 850,000
publications, respectively, from more than 450,000 authors. They are major services used by thousands of
computer scientists which provide fundamental support for scientists searching for publications or other scientists
in similar communities.
In this paper we describe a user friendly interface which plays the central role in searching authors and
publications and analysing social networks on the basis of bibliographical data. With the help of a sofisticated
combination of direct search and browsing through the data, the visualisation of social networks based on
co-authors, cited-by, conferences, etc., respectively, is the vital principle for finding publications which are
relevant to the user.
After introducing the concept of multi-mode social networks, the DBL-Browser itself and various methods for
multi-layered browsing through social networks are described.
Session: Friendship networks
Keywords: co-auhorship network , co-citation networks , cognitive networks
Accepted: Yes

223 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Alden Klovdahl
Austrailan Natural University, Sociology, School of Social Sciences, Faculty of Arts, Canberra, ACT 0200,
Mark Williams
Authors: School of Public Health, University of Texas, Houston, Texas
John Atkinson
School of Public Health, University of Texas, Houston, Texas
Sandra Timpson
School of Public Health, University of Texas, Houston, Texas
Sex at the Boundaries: Networks among men who have sex with men (and some women) in relation to infectious
disease control
One of the key challenges of contemporary infectious disease control is to better understand points at which
pathogens cross over from one group to another, thereby expanding the spread of infection and presenting more
difficult control problems. Here the focus is on networks of men who have sex with men, in particular a subset of
these who are male sex workers. Over 400 interviews were carried out with these men in Houston Texas
between 2004 and 2006. The data were collected with a link-tracing sampling design. About 25% of those for
whom the relevant information was available reported being HIV positive. Almost half had female sex partners in
the 30 days prior to interview. The total number of network links in this population exceeded 4,500, the average
per participant being about 12 (median, 9), with varying numbers of additional unidentified customers. Preliminary
analyses are presented. That is, considered are the number and type of components in the networks observed,
some of their structural characteristics (e.g., density, average distance, centrality), the location of HIV in the
population, the relevance of this for pathogen spread and control, and so on. Implications for future research on
social networks and infectious disease control are discussed.
Session: Infectious Diseases and Social Networks
Keywords: drug use , hiv , men who have sex with men , network structure , sexually transmitted infection , visualization
Accepted: Yes

224 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Andreas Klärner
Max Planck Institute for Demographic Research Independent Research Group Culture of Reproduction
Konrad-Zuse Str. 1, 18057 Rostock, Germany
Title: Measuring and exploring social networks – a mixed methods approach to social network analysis
A recurring problem in social network research is that collected network data differs depending on the way the
researcher asks for network-partners and that each network generator leads to a different sector of the
respondent’s network.

Our research is focused on the influence of social networks on fertility decisions and based on interviews
collected in two German cities. Since little is known about the connections between social networks and fertility
decisions, we decided to profit from a qualitative design and to collect network data as open as possible with
narrative interviews. On the other hand we wanted to collect systematically comparable network data, which led
Abstract: us to use the network chart introduced by Kahn/Antonucci and a network grid. We asked respondents to name
“important” persons in their life, leaving it open to the respondent to define what “important” means.

In this paper we evaluate this approach for collecting network data and ask:
- What kinds of relationships are included in the networks we explored?
- Who do the respondents consider as “important”? With whom do they talk about “personal problems”? Whom
do they experience as “supportive”? Who is “not important” or even “problematic”?
- How well do the measures derived from the network chart and network grid represent the accounts on the
network given during the qualitative interview?
Session: Mixed Methods: Combining Qualitative and Quantitative Approaches
Keywords: family , fertility , mixed methods , methodology , qualitative network
Accepted: Yes

225 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Andrea Knecht
Authors: a.b.knecht@fss.uu.nl
ICS/Dept. of Sociology, University Utrecht, Utrecht, The Netherlands
Title: Friends and Problem Behavior and Attitude in Early Adolescence. A Study of Selection and Influence Effects
Adolescent friends often show similarity in problem behavior and attitudes. Problematic behaviors and attitudes
during adolescence are, for instance, alcohol use, delinquent behavior and unfavorable attitudes towards
school-related behavior. Selection processes, where adolescents tend to form and maintain friendships with
others who are similar, and influence processes, where adolescents adjust their behavior and attitudes to those
of their friends, lead both to a positive relation of adolescents’ and their friends’ behavior and attitude. This paper
aims at disentangling influence from selection processes for alcohol use, delinquency, and attitudes towards
Abstract: school-related behavior and determining the relative strength of the association between these attributes and
selection and influence processes. Results from a longitudinal study of 438 students in 16 classes using the
SIENA software imply that friendship selection depends on the level of delinquency and friends exert influence on
the attitude towards school-related behavior. Friendship selection is moreover governed by the preference to
have reciprocated, transitive friendship ties to others of the same gender, and the tendency for adolescents to
maintain friendships from primary school. We found that attitude towards school-related behavior affects
delinquent behavior and delinquent behavior affects alcohol use.
Session: Adolescent Friendship Networks
Keywords: actor-driven modeling , siena , social influence , selection , alcohol use , school attitude , adolescence
Accepted: Yes

226 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Laura M. Koehly
NHGRI/NIH/DHHS Social and Behavioral Research Branch, Bldg 31, Rm B1B37D, 31 Center Drive MSC 2073,
Bethesda, MD 20892-2073, USA
Natalia R. Kuhn
Authors: NHGRI/NIH/DHHS Social and Behavioral Research Branch, Bldg 31, Rm B1B37D, 31 Center Drive MSC 2073,
Bethesda, MD 20892-2073, USA
June A. Peters
NCI/NIH/DHHS Clinical Genetics Branch 6120 Executive Blvd, EPS 7026 Rockville, MD 20852-7231, USA
Jennifer Loud
NCI/NIH/DHHS Clinical Genetics Branch 6120 Executive Blvd, EPS 7026 Rockville, MD 20852-7231, USA
A Social Network Approach to Communal Coping: Sisters at Risk of Developing Hereditary Breast and Ovarian
Communal Coping Theory offers a promising framework for investigating the social context within which families
cope with a shared stressor. Communal coping processes involve family members communicating about and
engaging in cooperative strategies to cope with a shared threat. This approach represents a shift from viewing
coping as an individual-level phenomenon to an interdependent relational process. To date the literature
regarding communal coping has either been theoretical in nature, or qualitative in design. Social network
methods offer a quantitative framework that can be used to further our understanding about communal coping
processes within families. The current paper illustrates the utility of social network methods for investigating
communal coping among sisters at risk of developing hereditary breast and ovarian cancer (HBOC). Sixty-five
Abstract: sisters from 31 families with known BRCA1/2 mutations provided data on their informational, tangible, and
emotional support networks using the Colored Eco-Genetic Relationship Map (CEGRM), perceived risk, and
psychological distress. Overlap in the support networks of sisters from the same family was used to assess
communal coping concepts (i.e. mutuality in exchange, pooled resources). Hierarchical models were used to
understand how network-based indices were related to similarity in risk perceptions and psychological distress.
Results suggest that sisters exchange information about their cancer risk; information exchange is affected by
similar risk appraisals; and shared support resources are related to psychological distress. Social network
methods provide an innovative approach to developing an understanding of how sisters within HBOC families
cope with their shared cancer risk.
Session: social networks and health
Keywords: breast cancer , depression , family , personal networks , social support , anxiety , ovarian cancer
Accepted: Yes

227 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Tina Kogovšek
Authors: University of Ljubljana, Faculty of Social Sciences, Kardeljeva Pl. 5, 1000 Ljubljana, Slovenia
Valentina Hlebec
University of Ljubljana, Faculty of Social Sciences, Kardeljeva Pl. 5, 1000 Ljubljana, Slovenia
Title: Network Composition in Name Generator and Role Relationship Approach
Different approaches in social network measurement have different (dis)advantages. Name generator approach
on one hand provides very detailed information on network characteristics (e.g., size and composition). However,
it is very time and money consuming and it requires considerable effort from respondent, when it is applied in
self-administered mode or complex coordination between interviewer and respondent, when it is applied in
personal interviews (e.g. Kogovšek et al., 2002). On the other hand, role relationship approach is cheaper and
easier to administer, but gives less precise estimates of network characteristics. With the role-relationship
approach unique identification of persons is possible only for “unique” role-relations, such as the partner. With
other role-relations multiple actual persons cannot be distinguished (e.g., friends, children, siblings). If we regard
each possible role relation functionally, this approach poses no particular limitation. However, estimation of the
network composition, a frequent practice in social network analysis, is limited, since we do not poses information
of the number of children, siblings and so on.

Alternative measures of network composition based on role-relationship approach (Hlebec, Kogovšek, 2005) are
compared and validated against measures obtained by name generator approach in 2x2 experimental design
with repeated measurement. Data were collected on a quota sample (defined by gender and age distribution) of
208 respondents in two waves (in each wave with a different data collection method) by the students of the Socia
Network Analysis course at the Faculty of Social Sciences in Ljubljana in October and November 2006. The
interval between the two measurement waves was two weeks.
Session: Collecting Network Data
Keywords: methodology , social support , personal network methods
Accepted: Yes

228 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Bruce Kogut
Professor of Strategy Insead Business School Fontainebleau, France
Authors: Mariano Belinky
Dept. Llenguatges i Sistemes Informàtics Universitat Politècnica de Catalunya Barcelona, Spain
Pietro Urso
Insead Business School Fontainebleau, France
Title: venture capital syndication as an evolving weighted graph
The vast proportion of studies on social networks focus on unweighted
graphs. While focusing on binary relations has proven to be revealing
in many social contexts, social relationships often are marked by the
renewal and amplification of ties. We show that the unweighted graph of
venture capital syndication lack many of the properties of social
networks of analytical interest: power laws, homophily, and robustness.
To the contrary, treating the network as a weighted graph of repeated
ties recovers these processes. We then decompose the syndications along
the dimensions of industrial and geographic classifications to study the
oscillations of the graph over time in reference to business cyles and
Abstract: to technological breakthroughs. Consequently, the study is able to
marry the social, economic, and technological processes to explain the
evolutionary dynamics of repeated tie formation. The data are drawn
from the VentureExpert database for the period of 1960 to 2005, as well
as from the US Census. By using a variation on Snijder's Siena model,
we estimate parameters to social rules conditioned on economic and
geographic measures. In all, we show that there is a tradeoff between
trusted expertise of established partners and the search for diversity.
These results are further explored by a simulation to test the
robustness of the network structure to small perturbations in the
Session: Network Dynamics
Keywords: venture capital , weighted graphs , network evolution , statistical models
Accepted: Yes

229 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Hanna Kónya
Authors: konya_hanna@yahoo.com
Corvinus University Budapest, Department of Sociology, Budapest, Hungary
Title: The personal network of a rural-local elite, emphasizing the patron-client relations
My paper is a case study, made in an ethnically mixed village in Transylvania (Romania). It refers to the local
elite, especially to the, so called "Patron-Client" relations.
The research focuses on the following questions:
-the homogeneous or heterogeneous nature of the personal network and the interpretation of ethnicity
Abstract: -social political background of the patron-client relation
-how friendship affects ethnicity
-the functional/utility relations of the local elite
The datas I analyzed were collected by questionnaire, interviewing and analyzed with Ucinet, Egonet and
discourse analyze. Although it is only about 20 people (the local elite) is a new approach of elite research.
Session: Social Capital
Keywords: community structure , ethnic identity , personal networks , stratification (elites)
Accepted: Yes

230 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Simona Korenjak-Cerne
University of Ljubljana, Faculty of Economics, Kardeljeva ploscad 17, 1000 Ljubljana, Slovenia, Ljubljana,
Slovenia, Slovenia
Natasa Kejzar
University of Ljubljana, Faculty of Social Sciences, Kardeljeva ploscad 5, 1000 Ljubljana, Slovenia
Vladimir Batagelj
University of Ljubljana, Faculty of Mathematics and Physics, Jadranska 19, 1000 Ljubljana, Slovenia
Title: Temporal clustering in citation networks
Temporal citation indices try to quantify the importance of works
(scientific papers, patents etc.) in time. It is very useful to know,
which works are currently interesting to the community or mainly
influence the newly emerging works, which works have the tendency
to become important, and which are on the other hand loosing its
present position. The information, that is used to quantify this
"importance" is the number and the time of citations for a
particular work. However we do not know exactly, what are the main
temporal citation patterns.
Our research in this work focus on the patterns. Given the time
distributions of citations, we would like to find the main classes
(the temporal distribution shapes) to which a particular work
belongs. We make use of the adapted leaders method (Batagelj,
Korenjak-Cerne), which we apply on the U.S. patents network data
from 1980 to 1999. We show, that the basic version of the method
inevitably leads to revealing single high yearly peaks, which might
obscure other distribution shapes also present in the data.
We try to modify the method according to this observation and
present our latest results.
Session: Knowledge Networks
access to knowledge , citation network , clustering , dynamics on networks , information seeking behaviour ,
longitudinal , methodology
Accepted: Yes

231 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Nikolaos Korfiatis
Department of Informatics Copenhagen Business School, Informatics, Howitzvej 60, Copenhagen 2000,
Marios Poulos
Department of Archives and Library Sciences Ionian University Plateia Eleutherias Palaia Anaktora, Corfu,
George Bokos
Department of Archives and Library Sciences Ionian University Plateia Eleutherias Palaia Anaktora, Corfu,
Title: Analyzing Scientific Networks using "Conflict of Interest"
Conflict of interest can be defined as the non-compatible relationship of actors of the same network where
relational ties denote an authoritativeness relationship when it comes to exchange. However existing measures
for research evaluation (e.g. the Impact Factor) fail to address these conflicting relationships on the analysis of
the academic prominence of a research. In this paper we describe a measure of prominence based on both
co-citation analysis and analysis of conflicting relations between authors. We present results based on example
data from a bipartite network of authors/co-authors and editorial board members from a popular online database
of scientific abstracts and discuss the inputs of social network analysis into the design of research evaluation
Session: Academic Scientific Networks
Keywords: co-auhorship network , co-citation networks , egocentric networks
Accepted: Yes

232 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Siru Korkala
Authors: siru.korkala@helsinki.fi
University of Turku (FI), Department of Sociology
Title: Appearance of trust in regional, co-operative organizational networks
Successful cooperation of businesses, professional organizations and other relevant parties depends on mutual
trust. Cooperation generates also social capital. Purpose of this research project is to study various networks in
the context of trust and social capital. Networks are cooperative initiatives in Southern Finland. Networks are
financed jointly by the Finnish Ministry of Labour and the European Commission. My research seeks for answers
to the following questions:
• how wide is the network in question, and what kind of changes occur in networks’ structure during the project
• how trust and partnership are being built in networks
• what kind of phenomena typically appear during the formation of networks
Network analysis, which is used for describing networks’ fine-grained structure, is the main method in use in this

In the paper, the networks are being scrutinized in a context, which is based on putting concepts of systemic trust
and personal trust in contradiction. Systemic trust appears in relationships with abstract systems, like norms and
regulations created by networking initiatives’ for themselves. The density of the networks varies significantly, and
also other types of diversity can be found. Density, centralization and the existence of structural holes are
studied. The networks are in very different stages of development. It appears that the higher degree of density
Abstract: and activity is present in a network, the more trust is based on the systemic level. On the other hand, the less
dense the network is, the more trust applies at personal level. This conclusion is illustrated by the Type of
Trust-Activity –Matrix that I will present at the conference.

Social networks consist of strong and weak ties. In business networks, strong ties implies to mutual agreements,
which define certain terms, conditions and ways to interact. Weak ties develop through experience, joint efforts
and gaining trust. In most cases, weak ties may connect also with organizations outside the existing network,
hence the networks extend. However, the extensions of this kind require that the newcomers should trust
strongly in the network as a system, not only in some of its members. In the beginning of co-operation, common
practices are not yet settled and network’s qualitative development has not taken place yet, hence trust exists
mostly on personal level. Personal trust does exist also in dense networks, but it is not equally dominant: norms
and regulations maintain trust instead. According to the social network literature, it is easier for dense networks to
maintain trust, and maintaining norms require high density. As a conclusion, one can say that building systemic
trust is a major prerequisite for effective and lucrative co-operation in the networks. Two out of three networks
under study are evolving continuously, thanks to the proactive attitude of the members involved. Future steps will
include testing the usefulness of network visualization methods, such as multidimensional scaling whilst
monitoring changes of positions of members in the networks.
Session: Business Networks
Keywords: organizations , social capital , ucinet , trust in networks
Accepted: Yes

233 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Johan Koskinen
University of Melbourne, Department of Psychology, Department of Psychology, The Redmond Barry Building,
Parkville, Melbourne, Victoria 3010, Australia
Title: Fitting models to social networks with missing data
Social network analysis is not exempt from the problem of missing data but dealing with missing data is typically
complicated by the interdependency of observations. The observed part of data may be dependent on missing
data and ignoring a non-respondent may be tantamount to redefining the network boundary. Many data
structures can be thought of in terms of missing data: multiple reports on an unobserved network structure;
observations only for a part of a network; observed ties being manifestations of underlying, unobserved
categories; etc. The conceptualisation of missingness may be given by the structure of data but may also depend
on the focus of the study (c.p. inferring latent structure in cognitive social structures as opposed to comparing
consensus of reports). In this paper the focus is on performing inference for models for the ties in a social
network with a fixed set of nodes, when we are confronted by missing information for dyads, i.e. for some pairs of
actors we do not know whether there should be an edge (arc) or not. Researchers have pointed to the difficulty in
dealing with this type of missing data and the pitfalls of not dealing with it but there are few suggestions as to how
we should deal with this type of missing data. Taking a Bayesian approach we propose a Markov chain Monte
Carlo algorithm that, given a few assumptions regarding what causes observations to be missing, allows us to fit
an (curved) exponential random graph model to social network data with missing information.
Session: Missing Data
Keywords: ergm (p*) , missing data , bayesian inference
Accepted: Yes

234 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Helena Kovačič
Faculty of Social Sciences Centre for Policy Evaluation and Strategic Studies (CPESS) Kardeljeva pl. 5 Ljubljana
Title: Network Analysis of Knowledge Flows Across Public Sector Organizations
This paper uses social network analysis to examine knowledge transfer relationships in the context of EU
integration processes. Within the last wave of the EU enlargement, several programs supporting accession have
been carried out between EU member states and candidate countries. These programs were mainly oriented to
the alignment of the acquis. The purpose of the study is to understand this alignment process in more details
since the processes of the adoption and the implementation of the new institutions are not necessarily the same,
Abstract: several have to overcome huge difficulties and not all meet the desired ends. This study examines and compares
different aspects of the knowledge transfer process: actors’ attributes (attitudes, perceptions), action of
knowledge transfer and characteristics of social networks (position, structure). Based on complete network data
form two EU projects I examine how social structure characteristics affect knowledge transfer process. Detailed
analysis gives us an insight into the relationships between network characteristics and outcomes of these
Session: Inter-Organizational Networks
Keywords: knowledge transfer , organizations
Accepted: Yes

235 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ralph Kremer
Vienna University of Economics and Business Administration, Interdisciplinary Department of Management and
Organisational Behaviour private address: Blumenstraße 25, Munich, Bavaria 80331, Germany
Capital, Structure & Functioning of ego-centered networks - an investment and financial theoretical approach on
sociological grounds
This article introduces an approach to capture a greater variety of the manifold influencing factors that constitute
the real value and usefulness of social networks for an individual. Beyond its apparent applicability on
professional levels, the “open” conception as a model of thinking and understanding makes it also valuable in
every personal and interpersonal setting, enabling the inclusion of many new and hitherto neglected factors.

Based on fundamental financial and mathematical understandings the presented methods could be used to
quantify and compare the value of ego-centered networks, however the primary intent is to present a method to
grasp the functioning and influencing factors that compose the actual value of Ego’s network.

The ostensible contrast of these two exertions finds its justification and grounds in the works of Pierre Bourdieu
who also developed models of thinking which derived from financial considerations to face the complexity of
social structures.
Far beyond the possibilities of visual illustrations this approach is designed to provide an instrument to implement
an (in theory) unlimited number of influencing factors that are rooted in the ”social being man” (homo socialis).

To explain its functionality the article introduces some common (intensity, frequency) and more important some
uncommon influencing factors (derogation over time, influence of personality) to present its wide applicability.
The different forms of capital are as much considered as the different social fields in which networks can be of
assistance and importance.
On these grounds further applications are left to the user’s creativity and field of study.
Session: egocentric network
egocentric networks , social capital , network structure , personal networks , qualitative network , social
networks , financial & investment theory , functioning of social capital
Accepted: Yes

236 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Lothar Krempel
Authors: krempel@mpi-fg-koeln.mpg.de
Max Planck Institute for the Study of Societies, Paul Str. 3, Cologne, Germany 50676, Germany
Title: Germany INC 2004: Capital Ties and Directory Interlocks
Using Data of the German Monopoly Commission I compare captial ties among the largest 100 German
Abstract: companies with their directory interlocks. While captial interlocks continue to decrease as is evident from
previous coauthored work board director interlocks point to an ongoing coordination of the German capitalism
Session: Inter-Organizational Networks
Keywords: changes in networks , 2-mode , inter-organizational networks , interlocking directors , network change
Accepted: Yes

237 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dean Krikorian
GroupScope, Inc. 2505 Solano Rd. Shell Beach, CA 93449
Greg Ludwig
Authors: GroupScope, Inc. 2505 Solano Rd. Shell Beach, CA 93449
Brad Cohen
GroupScope, Inc. 2505 Solano Rd. Shell Beach, CA 93449
Mirit Shoham
University of California, Santa Barbara, Department of Communication Santa Barbara, CA 93106, USA
Title: LinkSViewer 1.2 Network Analytic Software: Visual Bipartite Applications
LinkSViewer is a visual network analysis software tool developed by GroupScope, Inc. The LinkSViewer 1.2
Web-based product illustrates and investigates Venture Capital networks in the Silicon Valley. Our methodology
operationalizes venture relational capital based on four variables: companies, investors, board members, and
management teams. This demo/paper illustrates five applications of the software: (1) defining network variables
and attributes; (2) search and filter strategies; (3) the duality of single-mode networks (Breiger, 1974); (4)
network visualization (Krikorian & Ludwig, 2003); and (5) the pseudo-tripartite network. Each application is
presented using real-world examples as pictorial vignettes. The result is a visual network system that proposes
Abstract: an alternative way to look at relational database information. Relevant research applies social network literature
on affiliation networks (Simmel, 1955), social circles (Kadushin, 1968); social capital (Burt, 1992); and board
interlocks (Mizruchi, 1996). Practical applications examine the cases of board and management team interlocks,
IPOs, mergers and acquisitions, and corporate turnover, particularly from a relational venture capital perspective.
Implications involve the practice and pitfalls of designing Web-based network software, managerial uses of visual
network analysis, and online social networking applications. In the end we hope to showcase a real-world
application that yields insight toward further development in visual bipartite networks, business network
applications, and Web 2.0 software design.
Session: Visualization
Keywords: venture capital , inter-organizational networks , bipartite networks , visualization
Accepted: Yes

238 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Per Kropp
IAB Sachsen - Regional Research Network of the Institute for Employment Research Paracelsusstraße 12,
D-09114 Chemnitz, Germany
Title: Analysis of Regional Mobility Using Methods of Social Network Analysis (SNA)
Methods of Social Network Analysis are used to visualize and identify regional mobility patterns and labour
market regions in Germany. The results are compared with regional classifications usually employed in regional
labour market research. Furthermore, labour market regions themselves will be analysed and compared with
methods of Social Network Analysis providing a new empirical base line for regional studies.
Session: Networks, Economics, and Markets
algorithm , clustering , core-periphery , data representation , exploratory sna , germany , mobility ,
Accepted: Yes

239 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Aleksandra Krotoski
University of Surrey, Psychology, University of Surrey, Guildford, Surrey GU2 7XH, UK
Authors: Dr Julie Barnett
University of Surrey, Department of Psychology, Guildford, UK
Dr Evanthia Lyons
University of Surrey, Department of Psychology, Guildford, UK
Title: Making e-friends and influencing people: assessing the perceptions of opinion leaders in an online community
The 3D virtual world Second Life is a social space in which information exchange via formal and informal groups,
friendship networks and official channels has characterised the development of hierarchies of digital identity with
different access to knowledge and other assets. In this explicitly trade, barter and asset-accumulation
environment, the distribution of information capital has implications for economic cohesion and successful
community integration.

Using a five-degree snowball sample, 750 Second Life users responded to sociometric online surveys. This
paper explores the structural make-up of the resulting 6,849-node network, with the aim of theorising how
innovation is diffused through its population. This paper focuses on the opinion leaders in Second Life, and
specifically on other participants’ perceptions of these key actors. The analysis explores social psychological
characteristics of network nodes, with an emphasis on perceived trust, credibility and group prototypicality which
strongly emerge as important factors in online opinion leadership.

The paper also addresses assumptions about online network closeness, empirically assessing variations in
cohesion reported by respondents based upon the use of different modes of communication (public chat, private
chat and non-Second Life chat). By investigating the networks of one pervasive and online-only internet
community, this research extends existing online social network theory by focussing on pseudonoymous
interactions in the virtual world space.
Session: On-Line Communities
attributes , central actor characteistics , diffusion , identity , internet , perceived social networks , social
influence , trust
Accepted: Yes

240 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Yufu Kuwashima
Authors: yufu@jd5.so-net.ne.jp
Tokyo University, 2-9-7 Kikuna Kohoku-ku Yokohama-city, Yokohama 2220011, Japan
Title: The word of mouth effect in the network
I analyzed the relation between the word of mouth effect on consumer’s buying behavior and personal networks
in which consumers are embedded. Inmarketingresearch, the word of mouth effect is thought to be an outcome
of social interaction among consumers and it is argued to have positive effect on consumers’ buying behavior.
However, I argue that the effect would work differently depending on the nature of goods. I invested 65
consumers’ personal networks and their possession of conspicuous goods.(i.e. bags, wallets and shoes).I found
that the snob effect works for possession of goods with conspicuousness among cohesive consumers and that
the band wagon effect works forpossessionof goods with conspicuousness among structurally equivalent
consumers who are not cohesive. I also analyzed 100 consumers’ personal networks and possession of variety
of cosmetics. I found the band wagon effect works for possession of goods without conspicuousness among
cohesive and structurally equivalent consumers. Observed are the positive and negative effects of word of mouth
on consumers’ possession of goods. My findings suggest that the combination of consumers’ position within
personal networks and types of goods produces the difference in the word of mouth effect.
Session: Friendship networks
Keywords: 2-mode , blockmodeling , diffusion
Accepted: Yes

241 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Seok-Woo Kwon
Martin School of Public Policy and Administration 427 Patterson Office Tower University of Kentucky Lexington,
KY 40506-0027
Authors: Dan Brass
School of Management Gatton Business and Economics Bldg. University of Kentucky Lexington, KY 40506-0034
John Hagedoorn
MERIT and Department of Organization and Strategy Universiteit Maastricht P.O. Box 616 6200 MD Maastricht
The Netherlands
Title: The role of national trust in R&D alliances among European firms
Inter-firm alliances are a common form of organizational network in the strategic management research. In this
paper, we examine whether intercultural context between two network partner firms influences the governance
form of the alliance network: more specifically, our hypothesis investigates a hypothesis that the more national
trust a firm has towards its partnering firm's country, the less hierarchical governance structure is used to
organize its international alliance. To examine this idea, we analyzed a sample of 618 bilateral international R&D
partnership network, taken from the MERIT-CATI databank. These 618 partnership networks were sponsored by
542 companies from 12 countries, covering the period 1970-2003.

The dependent variable represents the choice of the governance structure for each R&D partnership network,
Abstract: and is coded 1 if the partnership network is organized as an equity joint venture and 0 if the partnership network
is organized as a contractual partnership. Our key explanatory variable, national trust, is taken from
Eurobarometer surveys. In order to show that national trust influences partnership governance form independent
of other notable factors, we included a variety of control variables on economic, cultural, institutional, and
geographical aspects of the countries involved.

Our empirical result suggests that national trust influences the governance choices of organizational alliance
network significantly. This implies that organizational alliance network is not just a dyadic event between two
organizations, but also shaped by the broad intercultural context, especially trust that people have about the
partnering country.
Session: Inter-Organizational Networks
Keywords: inter-organizational
Accepted: Yes

242 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Joe Labianca
INSEAD, Boulevard de Constance, Fontainebleau 77300, France
Andrew Shipilov
Authors: Gatton Endowed Associate Professor of Management Gatton College of Business & Economics University of
Kentucky Lexington, KY 40506
Valentyn Kalnysh
Ukrainian Institute of Public Health Kyiv, Ukraine
Yuri Kalnysh
Title: Career-related network building behaviors, range social capital, and career outcomes
We studied 459 social servants to determine how their career network building behaviors were related to their
social capital, and how this affected their speed of promotion within their organizational hierarchy. Results
suggest that greater use of network building behaviors leads to greater range social capital, which leads to faster
Abstract: promotion. Factor analysis suggested two main strategies for network building behaviors: networking within
structured groups organized around activities (foci) and informal unstructured networking. Informal unstructured
networking behaviors were related to greater range social capital and faster promotion, while structured foci
networking behaviors were unrelated.
Session: Social Capital
Keywords: social network , network surveys , inter-individual
Accepted: Yes

243 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Gina Lai
Authors: ginalai@hkbu.edu.hk
Department of Sociology Hong Kong Baptist University
Title: Shared Social Life and Access to Social Capital among Married Couples
The present paper aims to examine the social factors contributing to shared social life among married individuals
and the impacts on the access to social capital. Marriage unites two personal networks rather than merely two
individuals. The couple identity further facilitates as well as necessitates joint participation of married individuals
in social activities. Shared social life has been found to promote marital satisfaction and stability. However, little is
known about the contributing factors for this network dimension of dyadic relationship. Moreover, an extensive
social network would presumably bring rich resources for instrumental actions. Would shared social life limit or
enhance the opportunity for cultivating ties beyond one’s social circle, influencing the repertoire of social capital?
This issue warrants investigation. The analysis will be based on survey data collected from 801 Chinese married
couples in Hong Kong. Joint participation in activities undertaken in three contexts will be examined, including kin
network, friendship network, and community. The variations in participation by couples’ socioeconomic positions
will also be taken into consideration. Social capital is conceptualized in terms of network diversity and measured
by Lin’s position generator. Preliminary findings indicate that parents and better-educated individuals are more
likely to participate in family and friendship gatherings with their spouses. Respondents generally have a low leve
of participation in community activities and no significance variation is found across different socioeconomic
groups. Joint participation in social activities tends to be positively related to social capital accessed by
Session: Social Capital
Keywords: social capital , family
Accepted: Yes

244 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Josef Langer
Authors: josef.langer@uni-klu.ac.at
University of Klagenfurt Department of Sociology Klagenfurt Austria
Title: Cross-Border Networking of Business Organisations in theAlps-Adriatic Region - How is it in Southern Austria?
Since a number of years network analysis is attracting growing attention in management studies.
Cross-border networks are of particular concern, because they have to overcome a barrier which is not from the
economic but from the political sphere and as such beyond the control of the enterprise. European integration as
well as globalization challenge particularly Small and Medium Sized Enterprises in this respect. Border regions
are expected to get rid of their status as peripheries through cross-border relations. Small and Medium Sized
Enterprises are considered to be strategic actors in this process.
This contributions will present results of an empirical investigation of 70 companies in southern Austria. They
were part of a sample of 200 organisation including political, social and cultural units. All these organisations had
more or less strong ties to the Alps-Adriatic region outside Austria. It will be shown why and how these networks
were established and what role the border still plays in the relationship. Besides that the results give some hints
about centre of developments in this region. A new division of labour is emerging. Cross-border relations are vital
to participate in this development.
Session: Business Networks
actor agency , business networks , core-periphery , dependence centrality , embeddedness ,
entrepreneurship , network node
Accepted: Yes

245 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Barbara Lawrence
Authors: UCLA Anderson Graduate School of Management Box 951481, B506 Los Angeles, CA 90095-1481
Neha Shah
UCLA Anderson Graduate School of Management Box 951481 Los Angeles, CA 90095-1481
Title: Homophily: Meaning and Measures
Homophily is the tendency of individuals to associate with similar others, where similar others are those with
common attributes or values (Lazarsfeld & Merton, 1954). Research consistently finds that people select friends
and get advice from others of similar gender, age and attitudes, making homophilous behavior an uncontested
social phenomenon. Although McPherson et al. (2001) identify types, dimensions and sources of homophily, the
network literature has paid little attention to homophily theory. This paper presents an inductive analysis of 79
empirical articles that explores the core tenets of Lazarsfeld and Merton’s definition. What is a tendency, what
are associations and what makes one person similar to others? For instance, scholars define “tendency” using
two theoretically-distinct approaches: tendency as a deviation from expected values and tendency as a relative
difference from others. Espelage, Holt, & Henkel (2003) use the former. They find homophily in middle-school
peer groups because boys who fight reside within the same peer groups more frequently than would be expected
if they appeared randomly across groups. Eyal and Rubin (2003) use the latter. They assess homophily by
asking “strongly disagree--strongly agree” questions about television personalities, such as “The character
shares my beliefs”, and conclude that the lower the disagreement, the higher the homophily. Both authors
observe homophily, yet offer disparate explanations for why the association exists. We explore these and other
differences in meaning and measures by analyzing extant definitions of “tendency”, “associations” and “similarity”
and employ the results to begin developing a theory of homophily.
Session: Network Theory
Keywords: homophily , inductive theory-building
Accepted: Yes

246 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Emmanuel Lazega
University of Paris IX - Dauphine, Cerso - Sociology, Place du Maréchal de Lattre de Tassigny, 75775 Paris
Cedex 16, France
Lise Mounier
Marie-Thérèse Jourda
Title: Catching up with Big Fish in the Big Pond? Multi-level network analysis through linked design
This paper explores multi-level networks that observe two systems of superposed and partially interlocked
interdependencies, the first being inter-organizational, the second inter-individual. We propose the method of
structural linked design as a mode of articulation for these levels. This method is composed of two major steps:
first, we separately examine the complete networks at the two different levels. Second, we articulate the two
networks in relation to one another using systematic information about the membership of each individual in the
first network (inter-individual) to one of the organizations in the second network (inter-organizational). This
dual-positioning, or the linked design approach, is carried out in an empirical study examining performance
variations within the “elite” of French cancer research in 1999. By looking at measures of centrality, we identify
the actors that these top researchers consider as central or peripheral at the inter-individual level (the big and the
little fish among the elite), and the laboratories that the research directors consider as central or peripheral at the
inter-organizational level (the big and the little ponds among all the laboratories conducting cancer research in
France at that time). In addition to the rather trivial report of the competitive advantage of big fish in big ponds
(particularly because of the advantage of size for laboratories in this domain of research), we use measurements
of scientific performance to identify “catching up” strategies that the little fish use in this system. We suggest that
this method offers new insights into the multi-level dimension of complex systems of interdependencies, and also
into the way in which actors manage these interdependencies. We believe that this understanding adds a new
dimension to the sociological exploration of the determinants of performance, of meso-level phenomena such as
institutional change, or macro-level phenomena such as social inequalities.
Session: Multi-level networks
Keywords: multi-level networks , scientists , inter-organizational , inter-individual , linked design
Accepted: Yes

247 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

David Lazer
Harvard University, Kennedy School of Government, 79 J. F. Kennedy Street, Cambridge, MA 02138, USA
Authors: Ines Mergel
Harvard University, Kennedy School of Government, 79 J. F. Kennedy Street, Cambridge, MA 02138, USA
Allan Friedman
Harvard University, Kennedy School of Government, 79 J. F. Kennedy Street, Cambridge, MA 02138, USA
Title: Co-citation of prominent social network articles in sociology journals: The evolving canon
Social network analysis has been a particularly hot area across the social (and some non-social) sciences. How
has this growth, in turn, affected the field of social network analysis within sociology, the field which has served
as the primary home of social network analysis over the last several decades? In order to evaluate this question,
we examined the citation patterns of the social network papers in the two leading general sociology journals, the
American Sociological Review and the American Journal of Sociology, from 1990-2005, focusing on the body of
Abstract: literature that was cited by at least two papers in a given year. In short, we produced three network snapshots of
the social network canon during this period. These analyses reveal a combination of great change and
substantial continuity. There was a substantial increase in interest in social networks in sociology during this
period, and, in particular, an enormous rise in interest in small world issues, coupled with the abrupt entry of
mathematicians and physicists into the sociology canon. We will present the results of our co-citation analysis in
detail at the conference.
Session: Academic Scientific Networks
Keywords: affiliation networks , changes in networks , co-citation networks
Accepted: Yes

248 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Luis Leal
Authors: jleal@eui.eu
European University Institute, Department of Political and Social Sciences, Florence, Italy
Title: International Venture Capita Networks
Venture capital (VC) investing has become the subject of great interest because: i) it has played a unique role in
the development of high-tech industries in the United States, and ii) few European governments have been
successful in the design of VC programs aimed at creating a self-sustaining indigenous VC industry. While recen
results suggest that the presence of international VC firms can be beneficial to export-oriented startups, and that
foreign VC firms usually co-invest in partnership with local venture capitalists, no comparative study of
international VC networks has been undertaken to date.

Abstract: In this paper, we apply social network analysis methods to map out the evolution of syndication-based VC
networks associated to Canada, Israel, United States, and 10 Western European countries from 1995 to 2003.
Such networks describe co-investments made by local and foreign VC firms in indigenous high-tech ventures.
Our results emphasize the high-volatility of such networks, and that only specific network geometries are
self-sustainable and lead ultimately to a steady inflow of VC funds. Furthermore, our results shed light into the
processes through which new foreign VC firms are brought into the network and stress the role played by
brokerage and geographic proximity. These results are of interest to both practitioners and policy-makers
involved in the design of VC programs.
Session: Business Networks
Keywords: venture capital , business networks , international networks , entrepreneurship , high-technology
Accepted: Yes

249 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ho-Ching Lee
Authors: hoching.riverclear@gmail.com
Center of General Education Chung-Yuan Christian University Chung-Li, Taiwan
Title: Policy and Institutional Networks Matter: Understanding Climate Change Policy in Taiwan
Economic progress in Taiwan has led to greater public awareness of environmental protection affairs. Upgraded
from the Environmental Health Bureau, Taiwan’s EPA (Environmental Protection Administration) was established
in 1987 as the cabinet-level lead agency in environmental protection. Since then, policy measures have been
linked to a comprehensive regulatory system covering air, water, soil, solid waste and toxic substance control
and management. But in the case of climate change, contrary to its regulatory regime, EPA has only played a
marginal role.

Taiwan has ranked the 22nd, or one percent of the global emissions. Climate change and its subsequent impacts
also have diplomatic implications for Taiwan. Given Taiwan’s “unique” international status—a political, economic
and emission entity yet not represented in the United Nations, we now have formal diplomatic ties with 25
countries, twelve of which are island states, vulnerable to and likely to be adversely impacted by the threat of sea
level rise.
At its core, the formation and implementation of climate change policy in Taiwan can be better understood by
examining institutional and policy networks at play. These networks are usually referred as clusters of rights,
rules, incentive structures, decision-making procedures. They have helped explain how Taiwan’s governmental
actors like EPA interact with other governmental and non-governmental actors, characterizing the many informal
connections that link stakeholders together.

By adopting the analytical framework of policy and institutional networks, this paper attempts to discuss the
existing policy linkages between climate and non-climate policy issues at the international level; to examine the
cause of and institutional response to policy linkage at the domestic level; and at the interactive level between
domestic and international settings, to analyze to what extent the response to international policy linkage could
be explained by the domestic response following the logic of Putnam’s “two-level game.”
Session: Policy Networks on Climate Change
Keywords: climate change policy , kyoto protocol , adaptative governance
Accepted: Undetermined

250 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Seungyoon Lee
University of Southern California Annenberg School for Communication Los Angeles, California, USA
Authors: Bettina M. R. Heiss
University of Southern California Annenberg School for Communication Los Angeles, California, USA
Lu Tang
University of Southern California Annenberg School for Communication Los Angeles, California, USA
Transactional Versus Interactive Knowledge Networks: A Comparison of Interorganizational Network Topologies
in Biotechnology
Tie multiplexity has increasingly been explored as a fundamental aspect of social networks. Interorganizational
network studies have highlighted this perspective, suggesting that multiple types of ties interdependently
influence network evolution. This idea naturally leads to the joint consideration of network “flows” and network
“topologies” (Borgatti & Foster, 2003). Different types of content flowing through the network drive differential
motivation behind the network formation processes, subsequently creating dissimilar structural signatures at the
dyad, triad, and global levels. This study explores the topologies of two different knowledge networks in the
biomedical field: the transactional knowledge network (represented by licensing, marketing, and
commercialization alliances) and the interactive knowledge network (represented by research and development
Abstract: alliances). The network data are created from a recorded list of deals and alliances between 1989 and 2006. A
resource dependence theory perspective underlies the network formation process. The transactional knowledge
network operates as a short-term channel for the transfer of migratory and codified knowledge and access to
financial and marketing resources, e.g. large multinational pharmaceutical companies. Therefore, a star topology
with within-cluster centralization and high centralization at the global level is hypothesized. On the other hand,
the interactive knowledge network operates as a channel for the sharing of embedded and tacit knowledge over
time, e.g. across commercial and academic communities. Consequently, it is hypothesized that there will be
cohesive local clusterings and low centralization at the global level. Implications for the interaction between
network content and structure in multiple networks are discussed.
Session: Inter-Organizational Networks
Keywords: ergm (p*) , inter-organizational networks , knowledge networks , network structure , public-private partnerships
Accepted: Yes

251 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sin Yeung LEE

Authors: DSO National Laboratories, Singapore
Tong Kiat TAN
DSO National Laboratories, Singapore
Title: The Email White-Hole Effect
For a decision maker in an organization, we hypothesize that the set of email recipients is aligned more to meet
the demands of his work. A larger proportion of the emails he sends is meant for his immediate
supervisors/subordinates. He will persist in sending emails to these recipients even when some of them rarely
respond to him. Based on this conjecture, we want to ferret out the decision makers using email flows alone.

A user, u, is considered to be a white hole when a significant portion of the emails he sends gets no replies at all.
Note that a white hole is not necessarily a spammer since he does not mass mail indiscriminately. We define the
white-hole degree of u as the fraction of unreplied emails that u sent.
Using a one-month email log (97,000 email messages in total) of an organization with a population of 2775 users
we found that the average white-hole degree of the 86 decision makers was higher (0.5571) as compared to the
overall average white-hole degree of all users (0.3630). Furthermore, this degree was even lower (0.3486) for
lower echelon support staffs who played no role in the decision making process.

Similarly, we can define a black hole as a user who does not reply to a significant portion of the emails he
receives. We did not find any statistical difference for the black-hole degree for decision and non-decision
Session: Leadership Networks
Keywords: leadership and networks , organizations
Accepted: Yes

252 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Zong-Rong Lee
Authors: zrlee00@gmail.com
Academia Sinica, Institute of Sociology, 128 Academia Sinica Rd., Taipei, Taiwan
Title: Family Business and Formation of Social Capital
Family-controlled business involves specific kind of ownership arrangement and management practice. Its wide
prevalence and persistence have been reported both in Western economies and in emerging markets. While
current research has started to reveal its beneficial features and resilient adaptability in various aspects, little is
explored about how the organization characteristics of family-controlled business may be related to formation of
its network resources and strategic linkages. In this paper, I argue that family-controlled business is
endogenously favorable to establishing and accumulating general resources of social capital. I substantiate this
argument through the development of a theoretical framework that looks at three distinct dimensions of family
business that I consider generative to social capital: authority structure, temporal dimension of succession, and
kinship networks. This conceptual framework is operationalized and tested with data of intercorporate networks in
Taiwan where familism is deemed as one cardinal rule underlying most of business practices. I report results
from examinations on interlocking directorates, board ties of the chair, and equity alliance networks at firm and
business-group level. Implications for literature on social capital and family business research are discussed.
Session: Politics and Interlocking Directorates
Keywords: social capital , interlocking directors , family , kinship network
Accepted: Yes

253 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Pierre-Majorique Leger
HEC Montreal, Dept. Information Technologies, Montreal, Canada
Jean-François Dubois
Authors: HEC Montreal, Dept. Information Technologies, Montreal, Canada
Gilbert Babin
HEC Montreal, Dept. Information Technologies, Montreal, Canada
Robert Pellerin
Ecole Polytechnique de Montreal Dept. Mathematics and Industrial Engineering Montreal, Canada
The Longitudinal Diffusion of Enterprise Resource Planning Systems in Industrial Networks : The Case of the
North American Automotive Industry
Enterprise Ressource Planning (ERP) systems are commercial software packages that enable the integration of
transaction-oriented data and business processes throughout an organization. ERP systems display significant
network externalities as their usefulness depends on the number of adopters. This paper investigates the
longitudinal diffusion of enterprise system in business networks. Using public data sources, network data have
been collected for all years between 1994 and 2005 for the 50 largest firms in the auto sector in North America.
Firm-specific data related to the adoption of enterprise system was also collected. Specifically, we gather
evidence on the software adopted by each firm and the year of adoption of this software package. One specificity
of this research is that its presumes that firms may adopt more than one module from different software
packages and that firms may choose to change to another software provider over time. We present results from
the first phase of this research project. Only the four most important software providers in the field of ERP
systems were considered in the analysis. Preliminary analysis reveals evidence of an autocorrelation in the
adoption pattern and an inter-year propagation of the leading technologies.
Session: Innovation
inter-organizational networks , interlocking directors , diffusion , innovation , information technologies , supply
chain network
Accepted: Yes

254 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Juergen Lerner
Authors: University of Konstanz, Department of Computer & Information Science, Konstanz 78457, Germany
Ulrik Brandes
University of Konstanz, Department of Computer & Information Science, Konstanz 78457, Germany
Title: Comparing Networks by their Group Structure
A partition of the actor set of a network yields a smaller graph whose nodes are the actor classes and whose
edges summarize the ties between the classes. These networks on the class-level provide several ways to
compare different networks analytically and visually. Furthermore, they allow for comparing average
group-structures of sets of networks. In this talk we review several ways to obtain an actor partition, define the
class-level network, and analyze/visualize the resulting group structures.
In particular, we apply these methods to a set of approximately 500 personal networks of immigrants to Spain
and the USA which have been collected by Molina et al. Our visualizations make it possible to easily spot
patterns and outliers in hundreds of different networks, thereby providing insights into different types of
acculturation. For instance, average group-structures differ significantly between ethnic groups and are
dependent on time of residence in the host country.
Session: Mathematical Models
Keywords: methods , network comparison , acculturation
Accepted: Yes

255 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Laszlo Letenyei
Authors: llet@bkae.hu
U. Corvinus of Budapest, Sociology Dept, Fovam Ter 8, Budapest H-1093, Hungary
Title: Network Visualization of Mental Spaces with the Mental Map Editor
‘Mental map’ refers to our vision about the space around us. There are some network-based questions on the
topic as for example: the wealth or mental distance among mental spaces, the cognitive texture of a city etc.

Mental Map Editor is a software to

- prepare a questionnaire,
- realize a computer assisted survey asking for mental maps
- manage both relational, analytic and spatial data and finally
- analyze the data and prepare the report, by only a push of a button

Relational data gathered with the MME can be exported to Ucinet, Pajek or other SNA software, but it have its
own capacity too to collect and analyze relational data It have its own module for the visual interpretation of the
results. See more: www.mentalmap.org
Session: Visualization
Keywords: mental maps , visualization
Accepted: Yes

256 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ronit Levy-Zilberstein
Authors: The Hebrew University of Jerusalem Department of Sociology and Anthropology Jerusalem, Israel
Dr. Yuval Kalish
second author
Title: Are the Social Sciences Networking? An Examination of Citation Networks between the Social Sciences
This study set out to examine the potential discourse between disciplines of the social sciences. It was assumed
that the disciplines cite one another since all the fields examined are clustered under the umbrella of social
sciences. Since citations are a key instrument in academic research and evaluation of quality, citations analysis
was used to investigate the relations between these disciplines.

This present exploratory study examines the citation networks between 13 disciplines of the social sciences.
Citation data was collected using the ISI Web of Knowledge from 30 articles per discipline from top impact-factor
journals. Data including 9,275 citations which was then analyzed using Ucinet software.

Results show high variability between disciplines. For example, articles from economic journals often cite other
economic articles, a characteristic that types it as a "closed discipline". Sociology, by contrast, is central in the
citation network: most disciplines cite Sociology, and Sociology cites other disciplines. A blockmodel data
analysis revealed five main structurally-equivalent blocks. These five clusters were interpreted as a Political
cluster (comprising Economics, Political Science, International Relations, and Communication); a Social cluster
(comprising Sociology, Demography and Criminology); a Business cluster (comprising Business and
Management); an Individual cluster (comprising Psychology and Education) and an Isolates cluster (comprising
Anthropology and Geography).

This research highlights similarities and differences among disciplines belonging to the social sciences. The
results indicate a potential breakthrough in areas of the production and collection of knowledge.
Session: Knowledge Networks
Keywords: citation network , social sciences
Accepted: Yes

257 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Eleanor T. Lewis
Authors: eleanor_lewis@stanfordalumni.org
Center for Health Care Evaluation, 795 Willow Road, Menlo Park, CA 94025, USA
Title: Expertise, Availability, or Reciprocity? Explaining the Consultation Networks of Nurses in an Intensive Care Unit
Nurses in an intensive care unit must make immediate care-giving decisions that have observable and potentially
life-altering consequences for patients. Previous research suggests that information-seeking behavior in this
setting would be based on perceived expertise and availability. However, social network research has also found
a strong tendency in groups toward balanced relationships between members. In a group with an uneven
distribution of expertise but strong norms of cooperation and reciprocity, seeking information from others may
also depend on expectations of reciprocal behavior in difficult situations. This study examines the reported
interpersonal consultation networks of 32 nurses in an intensive care unit to explore competing explanations for
their information-seeking behaviors. Nurses did tend to report they would seek information from nurses they
perceived as having high levels of expertise. Nurses with an intermediate level of nursing experience, however,
reported differentiated information-seeking behavior with novice nurses, while nurses perceived as experts were
either highly selective or very unselective in who they reported they would consult with. There were also
meaningful differences between the day and night shifts in their reported consultation networks.
Session: Intra-Organizational Networks
information seeking behaviour , access to knowledge , informal networks , intraorganizational networks ,
perceived social networks , reciprocity , health care providers
Accepted: Yes

258 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Pengxiang Li
Xian Jiaotong University, Department of Management Science, School of Management, Xian Jiaotong University
Xianning west on the 28th, Xian, Shanxi Pronvince 710049, China
Title: The Scale-Free Behavior of the Hierarchical Structure Embedded in Social and Business organizations.
This paper explores whether there exists or not scale-free behavior in the hierarchical structure network popularly
embedded in social and business organizations. Theoretical analysis demonstrates that its accumulation degree
distribution follows the power law of P(K)=AK-2 [K=kacc+m/(m-1)] for the simple growth model with a constant
span of control m. We also find that the complex growth model with a random span of control satisfies the
conditions of growth and preferential attachment that may lead to scale-free network. However, another
mechanism of equal opportunity will put into action when two nodes have the same accumulation degree, which
Abstract: changes the power law distribution a little towards the Poisson distribution featured by random network.
Numerical simulation demonstrates that accumulation degree distributions for the above two models have an
exponential tail, the former with degree exponent of 2 and the coefficient A of m2/(m-1), the latter with a curve
head and the coefficient less than m2/(m-1). These findings show that the hierarchical structure network has
indeed the scale-free behavior. The research results about scale-free network are also applicable to the network
of organization structure. To discuss the implication of scale-free behavior for management science research is
not meaningless.
Session: Complexity
business networks , dynamics on networks , network evolution , network structure , social network , scale-free
Accepted: Yes

259 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Klaus Liepelt
University of Applied Sciences, Mittweida, Germany, Department of Media, Hochschule Mittweida,
Technikumplatz 17, Mittweida D-09648, Germany
Title: Shortcuts make the Difference. On the Emergence of Decision Clusters in Germany's Parliament
Abstract: The paper builds on previously presented research on informal inter-caucus relationships of German
MPs and their contribution to policy outcomes. It will follow up on the finding that success or failure of legislative
projects has to a considerable degree been reflected by divergent interaction patterns which prevail among about
20 decision units (Parliamentary Committees) constituting the legislative production line. Particular identities and
transaction styles within these policy production units have emerged and are continuously being reproduced.
Momentum and quality of this process depends on the existence of relatively small trans-partisan “nuclei of
competence” that find each other on political shortcuts.

These nuclei, drawing on the social capital of their frequently hidden members, serve to modulate conflicting
interests in their fields – be they ideological, partisan, or clientele-oriented - with the goal of consensual
outcomes. As each of the parliamentary production units pursues its policy outcomes in a quasi monopolistic
sector, competition between them is low, albeit for the financial resources to bestow their legislative domain. The
solutions they present bear the earmark of intra-committee compromise. The failures that occur may reflect an
inability to cope with demands from external forces. Certainly, getting consensus does not depend just on the
special qualities of the Hidden Few: It is easier to maintain a consensual stance in Foreign Affairs than to cope
with the conflicting interests of Health Insurance.

If salient issues lead to widespread conflict, any quality measure of policy production procedures looses weight.
Anyway, an attempt is being made to detect the competence nuclei for each of the 20 policy domains and to
analyze them in network terms. Empirical evidence for structure and action potential of these “nuclei of
competence” will be presented comparatively. It is expected that dissecting the global legislative record to its
proper decision levels will help to better understand the diverse interaction patterns of the parliamentary
processes. Their mix may be different from time to time, and it yet waits to be quantified. But these three have
always coexisted: Top-down government/opposition (“interface” disciplined), multi-level stratarchic (“arena”
disciplined), and bottom-up solution-oriented (“council” disciplined) components of the political discourse.

A post hoc case study of the “media faculty” contingent among the MPs will give an example of how a “nucleus of
competence” may emerge in a transformation process that reduces 34 possible “contacts” to a maximum of 10
that eventually feed a special “Channel”. It is then argued that the process toward emergence (as well as
disappearance) of such acting nuclei may best be observed in real time by analysing the public statements that
continuously produce the stories among interacting players. The “Transparent Parliament Research Group” at
Mittweida University has started a program to accompany the ongoing parliamentary Health Reform discourse
using process produced data: Daily news agency reports are now producing continuous Automap records of
Actors, Institutions, Events, Ressources. As the Actors not only include MPs but any other Actors that moves on
the health playground, it is expected that the “Hidden Few” that competently aim at controlling a dynamic and
complex process, will shape up. First results from this study will be presented to round up the argument.
Session: N/A
Keywords: political networks , policy network , dynamics on networks , germany , parliaments
Accepted: Yes

260 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jorge Lima
Authors: jlima@notes.uac.pt
Universidade Dos Acores, Dept. Educacao, Rua Da Mae De Deus, 9501-801, Ponta Delgada, Portugal
Title: Combining network and interview data: strenghts and challenges of a multimethod approach
Network methods provide powerful ways of looking at social structures and systems, but they can hide as much
as they can tell. Interviews provide alternative ways of understanding network processes. The combination of
Abstract: both research approaches can help gain more in-depth knowledge of actors' embedded social lives. This paper
discusses the strengths and shortcomings of these two approaches, as well as ways of combining them. Actual
examples of research applications of this combined, multimethod strategy are given and discussed.
Session: Qualitative Social Networks Research
Keywords: method triangulation , methodology , qualitative network
Accepted: Yes

261 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Ching-Yung Lin
IBM T. J. Watson Research Center, Hawthorne, NY, USA
Authors: Kate Ehrlich
IBM T. J. Watson Research Center, Cambridge, MA, USA
Vicky Griffiths-Fisher
IBM Global Business Service, London, UK
Title: Small Blue: Making Big Blue a Small World with social sensors, expertise search and social network analysis
The most valuable asset of a corporate is arguably the knowledge and experience of its employees. People
mostly rely on personally network of friends and colleagues to get trusted information, referral and collaboration.
According to a 2006 Gartner report, corporate employees get 50 - 75% of their information directly from other
people. But, in a global enterprise, e.g., IBM -- the Big Blue, we usually do not know where the right people are
and do not know how to reach them so they are willing to help. Finding through personal network is usually slow.
Searching through profiles is usually futile, because most people do not put their detailed skills and knowledge
into profiles, and gets obsolete information. Recently, manual social tagging was proposed as a solution to
facilitate expertise search. However, tagging is mostly static. Furthermore, studies showed that, unlike teenagers
corporate employees do not want to tag other employees. They either do not have time or do not want to do so.

SmallBlue is an IBM expertise search and social networking suite that automatically captures and visualizes
people's private personal network, helps to manage and expand it, and find the shortest path to access people
with specific knowledge or skills through personal extended network. It is also a public expertise search engine
and social network visualization and analysis tool for large corporations. We built a SmallBlue Social Sensor
system which deploys automatic sensors at volunteers' machines to gather their authored data for large-scale
dynamic social network and expertise inference. We provides three applications: SmallBlue Ego, SmallBlue Find
and SmallBlue Connect, for personal social capital management, expertise search/reach and social network
analysis, respectively. Two months after its internal release, the system has been installed at more than 1,200
volunteers' machines in 48 countries and has been dynamically collecting and inferring the expertise and social
network of more than 130,000 people in corporate. This project is swiftly adopted by internal business users. We
are working towards a public version for general organizations.

We will give an overview on the system architecture, artificial intelligence inference algorithms, user interfaces,
features in dynamic large-scale social networks, privacy and security issues, user studies, and demos.
Session: N/A
collaboration , content analysis , corporate social capital , dynamics on networks , informal networks ,
intraorganizational networks , knowledge network , personal networks
Accepted: Yes

262 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Nan Lin
Duke University, Department of SociologyBox 90088, Durham, NC 27708-0088, USA
Authors: Chih-jou Chen
Institute of Sociology Academia Sinica Taipei, Taiwan ROC
Yang-chih Fu
Institute of Sociology Academia Sinica Taipei, Taiwan ROC
Title: The Institutional Basis of Social Capital in Taiwan and PRC
The consequences of social capital for instrumental (e.g., job mobility) and expressive (e.g., sense of well-being)
outcomes have been well documented. In this paper, we examine the production of social capital – which
institutions in the two societies affect the distribution of social capital (the diversity or extensity of occupations one
may access in one’s social networks). Specifically, we are interested in how social, political, economic and
cultural institutions differentially affect the distribution of social capital in Taiwan and mainland China. Analyses of
Abstract: data from two national surveys conducted in Taiwan and urban China in 2004-2005 show that gender, education,
active participation in voluntary associations, and socioeconomic positions in both societies affect the extent of
one’s access to various occupations. Also, in mainland China, political affiliation (e.g., party membership) affects
one’s access to cadres, whereas in Taiwan, employment (i.e., in the public sector or self-employment) is
associated with more extensive social capital. We discuss these similar and different patterns in terms of the
different political and economic regimes.
Session: Social Capital
Keywords: social capital
Accepted: Yes

263 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

John Liu
Authors: Vanung University, Department of Business Administration, Chungli, Taoyuan, Taiwan
Chyan Yang
National Chiao Tung University, Institute of Business and Management, Taipei, Taiwan
Title: Herding in Affiliation Networks
We propose an affiliation network model that captures the herding behavior commonly observed in affiliation
networks. Herding is the phenomenon of the same two or more actors participating in the same two or more
events. Paraphrasing in the terminology of board of directors, herding is the multiple interlocks of directors, or,
the situation when a group of board directors who sit side-by-side on not just one but several company boards.
This local social process in the affiliation network is translated into multiple links in the projected one-mode
networks. From a global perspective, herding has the effect of lengthening the geodesic path of the one-mode
Our model is a simulation model based on the arbitrary degree distribution concept developed by Newman,
Strogatz and Watts (2001). The board degree distribution of a purely random affiliation network is known to
deviate from the real world board network of similar characteristics. We introduce the herding behavior into the
random generation scheme to balance the randomness with the intended social process. The method creates
many multiple interlocks and results in a better approximation to real world board networks. In applying the
model, several parameters highlighting the herding behavior such as bipartite clustering coefficient, (Robins and
Alexander, 2004), redundancy coefficient (Latapy et al., 2006), social inertia (Ramasco and Morris, 2006) are
compared against the real world network. We use board of directors network in Taiwan as the reference base for
our comparison.
Session: Complexity
Keywords: affiliation networks , interlocking directors , network structure
Accepted: Yes

264 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Li-Wen Liu
Authors: lwliu@mail.thu.edu.tw
Tunghai University, Department of Social Work PO Box 922, Taichung City, Taiwan
Title: Networks as Strategies of Coordination for Organizations in Mental Health Rehabilitation after Natural Disasters
In the past few decades, the natural disaster, including earthquake, flood, drought, hurricane, has taken place
everywhere around the world. It took millions of lives and caused enormous loss, including home damages,
financial losses, and business hardships. The issues regarding natural disasters have increasingly been the most
challenging aspect confronting human beings.

One of the most challenging aspects for organizations that serve victims of the disaster is the psychological
impact. It is important to note that providing aid for mental health rehabilitation to disaster victims requires diverse
types of services from various organizations. Therefore, developing a coordinated mental health service system
for people suffered from natural disasters has become an increasingly critical issue for service agencies in the
Abstract: communities affected.

Using network analytical techniques and data from a community in Taiwan, the author aims to examine patterns
and strategies of organizations in building a coordinated service delivery system for disaster victims. Data for this
research were collected at two points in time to compare overall network structures. It tends to describe changes
in the service network over time. In addition to interorganizational approach, both clients’ and orgnaizations’
perspectives on service availability, service quality, and service coordination will be examined in order to assess
system performance. Questions regarding the conditions of organizations that determine coordination in service
delivery is also be examined. GIS Mapping Techniques were used to display information regarding various
conditions, including damage assessment, organizational resource inventory, and scopes of service networks.
Session: Inter-Organizational Networks
coordination , health care providers , inter-organizational , network structure , public-private partnerships ,
changes in networks , natural disaster
Accepted: Yes

265 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Paulette Lloyd
Indiana University, Bloomington, Sociology Department 1020 E. Kirkwood Ave, Room 744 Bloomington, IN
Authors: 47405-7103, USA
Jan de Leeuw
University of California, Los Angeles Department of Statistics, Chair 8125 Math Sciences Bldg Box 951554 Los
Angeles, CA 90095-1554, USA
Title: Mapping Globalization: modeling large scale political and economic changes
This paper addresses competing theories of world divisions using social network analysis and a model we
describe as a geometric representation, combining aspects of log-linear analysis with multidimensional scaling
and correspondence analysis (“EDM” or exponential distance model) to analyze datasets on United Nations
General Assembly (“UNGA”) voting and world trade. We examine and compare changes in the structure of state
Abstract: alliances and world trade centered upon the decades before and after the end of the Cold War. Our model
incorporates external information (state attributes such as political regime and levels of economic development
and state IGO membership) to explain the map of voting. The model combines regression and spatial analysis in
order to explain the shifting bases of state alliances across multiple dimensions of social life—economic, political,
cultural/ideological and military.
Session: Political Networks
Keywords: social network , world systems , trade , political networks , globalization , economic networks
Accepted: Yes

266 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Laurence Lock Lee

Authors: locklee@iinet.net.au
Sydney University, Faculty of Economics and Business, Merewether Building H04, Sydney, NSW 2006, Australia
Title: Corporate Social Capital Links to Firm Performance in the Global Information Technology Services Sector
The key aim of this research was to explore the impact of Corporate Social Capital (CSC) on firm performance.
Through the use of CSC as a dominant lens for viewing a firm’s intangibles, a number of important
sub-components of the CSC formulation are exposed. These include a firm’s market centrality, absorptive
capacity, internal capital, human capital and financial soundness. Firm performance was measured as return on
investment (ROI), market to book ratios (Tobins Q) and total shareholder return (TSR).

Overall the research identified CSC as a significant predictor of firm performance, but fell short of fully explaining
the market to book value disparity. For this research, an innovative computer supported content analysis
Abstract: technique was devised to capture a majority of the data required for the empirical research, and to build the
alliances network representation of the global IT marketplace.

The research results for market centrality found that high centrality can indeed be a liability for the larger more
mature firms, but is definitely an asset for the smaller software firms. Financial soundness was the most
predictive of the CSC elements for firm performance, but interestingly the relationship was also weakest for large
mature firms. For small firms with poor finances, investment in human capital was found to be the only
performance enhancing investment. Investments in R&D and internal capital were only viewed positively by the
share market for firms who had the perceived financial capacity to support such investments.
Session: Business Networks
Keywords: business networks , inter-organizational , corporate social capital , content analysis , intellectual capital
Accepted: Yes

267 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Alessandro Lomi
Authors: University of Bologna, Department of Management Science - Faculty of Economics, Via Capo di Lucca 34,
Bologna 40126, Italy
Philippa E. Pattison
Title: What does it mean for you? An exploratory study of organizational identity across levels
A central problem in contemporary theories of organizations is how to represent the connection between systems
of action and systems of symbols – the two main productions of actual organizations. The essence of this
problem seems to be that neither the world of action nor the world of symbols can exist – or be structured –
independently. Rather, each can be understood only through association with the other. For this reason
theoretical approaches that either reduce organizational action exclusively to sets of material practices and
routines (or behaviors), or to cognitive processes of construction and maintenance of coherent systems of
meaning appear to be equally incomplete. These general considerations on the mutually constitutive character of
organizational symbols and practices orient the methodological proposal that we articulate in this paper. Using
lattice models we produce preliminary results from the analyses of a relational data set resulting from the
empirical association between 47 top managers in a multinational industrial group and 30 descriptors of
organizational and individual identity in form of words. The association between individuals and words is recorded
across three distinct – but hierarchically nested – social settings giving rise to a complex three-way connection
between individuals and their symbolic descriptions of identity. In the lattice representations that we propose
every point provides information both on the logical order that underlies the connection among individuals
through words, as well as the logical order that regulates the relation among words through individuals. Lattice
analysis reveals the classificatory structure of the data and allows us to interpret the underlying logic of the
duality of individuals and their symbolic representations (descriptors of organizational and individual identity in
this case)
Session: Intra-Organizational Networks
Keywords: affiliation networks , lattice models , organizations
Accepted: Yes

268 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Markku Lonkila
Authors: markku.lonkila@helsinki.fi
University of Helsinki, Dept of Sociology, PL 18 (Unioninkatu 35), 00014 University of Helsinki, Helsinki, Finland
Title: Antimilitary activism in the Russian internet
According to the European Values Study, army is one of the few institutions in which Russian citizens have a
reasonable amount of confidence. However, for several reasons such as the continuing tradition of hazing and
bullying in the army, many draftees try to avoid the military service seeking assistance among other things from
the various factions of Soldiers’ Mothers’ organizations.
This study examines Russian antimilitary activism through the analysis of the ‘issue-network’ of interconnected
antimilitary websites in the Russian-language part of the internet. What kind of topics are being debated? Who
are the main actors and how are they related? The data is collected with IssueCrawler, a software constructed to
produce a network graph of the interrelated websites.
Session: Collective Actions and Social Movements
Keywords: collective action , civil society , russia , internet , activism
Accepted: Yes

269 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Trond Loyning
Authors: trond.loyning@sos.uib.no
University of Bergen, Sociology, Rosenbergsgt. 39, Bergen 5015, Norway
Title: Mapping government – business ties in Norway: From inclusive policy networks to exclusive expert networks?
A number of studies show how governmental regulation of business has changed during the last decades. It has
been argued that networks are of increasing importance in new governance regimes, both internationally and at
the national level.
In this paper, ties between government and business in Norway between 1980 and 1995 are investigated. The
data consists of directors of the largest corporations and members of public committees. The latter has
traditionally been seen as an important part of Norwegian corporatism. It is shown that these networks are
reduced in size and density during this period. However, the changes are uneven: by analyzing various
subnetworks it is shown that this reduction appears among corporate boards and among public committees; the
number of ties between corporate boards and public committees are not reduced. In addition, analysis of a
particular group of actors, those related in one way or another to the financial sector, show an increasing number
of connections among these actors, resulting in increasingly central positions in the networks. The result is
discussed in relation to changes in the corporatist system in Norway; to changes in the institution of public
committees; and to changes in governmental regulation. It is argued that despite the overall trend in the network
patterns, networks appear to be of increasing importance in governmental regulation of business.
Session: Politics and Interlocking Directorates
Keywords: interlocking directors , governance , political networks
Accepted: Yes

270 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Xiaohui Lu
Authors: bizlxh@nus.edu.sg
National University of Singapore Department of Business Policy Singapore
Title: Network Extension: Creating Entrepreneurial Networks through Referrals
How do entrepreneurs employ their network relationships to promote new venture performance? Drawing on both
qualitative fieldwork and quantitative analysis of 100 high technology entrepreneurial firms in Zhongguancun
Science Park—China’s Silicon Valley, this study shows that entrepreneurs have used their network ties not only
for access to resources, but more importantly, for creation of new ties with potential resource holders. I call this
process network extension— expansion of an ego-centric network by addition of new ties resulting from
successful referrals made by existing contacts. I argue that network extension represents a key mechanism
through which networks influence new venture performance, especially in emerging economies characterized by
Abstract: high uncertainty and low trust. That is, more effective network extension leads to better entrepreneurial
performance. I then explore the antecedents of network extension. In other words, which entrepreneurs are more
effective in generating referrals from their existing network contacts? Specifically, I examine how tie strength,
entrepreneur’s prior employment, entrepreneur’s university alumni affiliation, and entrepreneur’s social absorptive
capacity impinge on network extension. Overall, this paper puts forward a theoretical framework which specifies
the network mechanisms that lie beneath the link between entrepreneurial networks and performance on the one
hand, and identifies the key factors that give rise to such networks on the other hand. Statistical analysis provides
support for the proposed theoretical model.
Session: Business Networks
Keywords: entrepreneurship , contact use , egocentric networks
Accepted: Yes

271 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Miranda Lubbers
Universitat Autonoma de Barcelona, Personal Networks Lab Edifici B-Facultat de Lletres 08193-Bellaterra,
Barcelona 08193, Spain
Jose Luis Molina
Universitat Autonoma de Barcelona Personal Networks Lab Edifici B-Facultat de Lletres, 08193-Bellaterra,
Barcelona 08193, Spain
Chris McCarty
University of Florida Survey Research Center Bureau of Economic and Business Research PO Box 117145
University of Florida Gainesville, FL 32611-7145
Title: A typology of personal networks of immigrants as an indicator of acculturation
The present proposal explores whether personal networks of immigrants differ as a function of their years of
residence in the host country. We assumed that time-related differences of the networks would reflect an
increasing acculturation in the host society. As such, we hypothesized that networks of recent immigrants would
be dense and homogeneous, and that the diversity of a network in terms of composition and structure (e.g.,
number of clusters in the network) would increase over time.
In order to test our model, data were used of 535 respondents of four immigrant communities in Barcelona and
four immigrant communities in New York. Computer assisted personal interviews were performed using Egonet.
Respondents were asked to free-list 45 alters, and to give additional information about these alters on a range of
variables (such as country of origin, country of living, strength of the tie, frequency of contact, role in ego´s
network). Moreover, they evaluated, for each of the 990 dyads, whether alters knew each other.
Abstract: On the basis of our general hypothesis, we expected that several structural and compositional characteristics of
personal networks would be related to the level of acculturation and hence would differentiate types of networks.
For the Spanish data, a k-means cluster analysis was performed on the basis of these characteristics, and it
indicated that a four-cluster solution was best interpretable. This implies that four types of personal networks
could be distinguished in the data, which are discussed in our paper. The four types indeed differed significantly
in years of residence (an external variable in the cluster analysis), in fact suggesting two pathways of network
change. Gender and education did not differentiate the types of networks, but country of origin did. However, the
relation between the type of network and years of residence was still significant when controlled for country of
origin. Analyses were repeated for the American data.
The paper discusses the use of the results for a longitudinal analysis, for which follow-up data are currently being
Session: Personal Network Methods
Keywords: migrants , personal networks , acculturation
Accepted: Yes

272 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Linda Ludwin
Authors: l.ludwin@chester.ac.uk
University of Chester, Arts & Media, Parkgate Road, Chester, Cheshire CH1 4BJ, England
Title: The Sticking Power of Social Capital: How Networks and Networking Enable UK Feature Film Production
Context: UK Feature Film Production

Harvard economist Richard Caves (2001, p. 103) called the way in which films are financed “one-shot deals
involving many specialists” to convey the temporary and limited nature of the involvement of all parties. Daskalak
and Blair (2002), who write about UK film units, explained: “This ambiguous and complex activity…[has] been
conducted in temporary organisations, with organisational structures and resources being constructed and
assembled on a recurrent but temporal basis” (p. 8). Sahlin-Andersson and Söderholm (2002) have described
projects (which they also refer to as temporary organizations) as “organized in terms of an effort that is running
from ‘unfinished’ to ‘finished’” (p. 20). Ekstedt (2002) has described temporary organizations as “having the three
t’s in common: time, task and team. Projects are mostly defined by the task they are supposed to fulfill [sic]
during a specific time” (p. 61).

Within a period of less than a year, each unit ranges from employing a single individual to employing many
hundreds of specialists, and uses an astounding range of sub-contractors for the provision of goods and
services. Unit numbers dramatically decrease during the period of time after filming, when editing and sound
work take place, dwindling to the handful of staff necessary to deliver the finished film and wind up the company.
Once a production ends and people take up their next jobs, individuals who worked together in a particular unit
are instantly dispersed. In terms of finance, at minimum - on a very low budget feature - upwards of a million
pounds is spent within the same brief time frame. In 2003, the median inward feature co-production budget was
£47 million; the median UK domestic feature budget was £3 million for the same year (UK Film Council Statistical
Yearbook 2003, Chapter 12, p. 3).

A unit’s business can take place anywhere in the world, and often does, sometimes on locations in many differen
countries during the course of an eight or ten week shoot. Feature films can involve any type of equipment or
physical setting imaginable. Every unit faces unique logistical, practical, financial and artistic demands generated
by its particular script.

These extreme temporary organizations, with their nomadic, self-employed personnel moving relentlessly from
one temporary organization to the next, often changing job descriptions along the way, may seem to represent
the ultimate in out-placement, contract culture and just-in-time philosophy. For example, in the Skillset Feature
Film Production Workforce Survey 2005, 51% of all freelances reported working on 2 or more films annually
Abstract: (supplemented by other audio-visual work); 57% had an involvement of 12 weeks or less; 70% had an average
working day exceeding 11 hours. In 2001, Blair, Grey & Randle reported on a unit where the average working
week for freelances was 69 hours over 5.6 days. 59% of the crew they studied had always worked freelance and
had never been employed, with entrants after 1990 predominantly always freelance.

On film units, most cast and crew members meet for the first time at a social occasion. The next day the
production is in full swing. How do people who work in temporary organizations know how to work together? How
does learning take place so rapidly and effectively in such transient settings? What happens to the learning in a
temporary organization when it ceases to exist? Is learning transferred from one temporary organization to
another, and if so, how?

Research question: Of what importance is social capital for temporary organizations such as UK Feature Film
Production Units?

Networking is a key activity in the UK’s feature film production sector. Through networking, the sector’s core
temporary organizations, film units, are brought into being. Networking enables units to function immediately.
Networking underpins the methods through which organizations, and the individuals of whom they are composed
work together. Networking enables learning and knowledge transfer. Networking has sustained the UK’s fragile,
complex and economically important film production sector for more than a century.

This paper considers how individuals freelancing for UK film units are both enabled and required to access
(better) social capital through networks and networking. It explores the centrality of social capital to film units and
to the sector as a whole. It explains how networking drives various film unit processes. It suggests special
features of film production units whereby human capital, cultural capital and social capital are structurally
interconnected so that all can be achieved simultaneously. The paper’s central question is: of what importance is
social capital for temporary organizations?

Research methodology: Case study approach

The central research question will be examined in the context of the UK’s feature film production sector, drawing
on material from an embedded case study completed by the author in 2004, The One-shot Deal: Temporary
Organizations, UK Feature Film Units, and Learning Organization Theory, a variety of contextualising secondary

273 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

274 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jar-Der Luo
Authors: jdluo@mail.tsinghua.edu.cn
Tsinghua University, Sociology, #33 Lane 1 Sec.4 Hsing-Lung Rd., Beijing, Beijing 100084, P.R.C.
Title: How Small Groups Form in an Organization -- A Simulation Model for the Distribution of Working Load
This paper will examine the reasons why an organization’s employees form small groups by simulating their
networking behaviors. Let us assume there is a department in an organization with 20 employees. All of them
execute very similar job, so that they can mutually support each other. Whenever any person can’t finish his/her
job in one day, he/she will ask colleagues for help. Our model illustrates that mutual help make people tie
together, and some of them form mutual support association due to the reciprocity principle and efficiency of
resource exchange. Eventually, these associations exclude others’ participation. That is a reason why small
groups form in a firm. Our simulation will explore what kinds of working environment turn mutual-help behaviors
to exclusive mutual-support small groups.
Session: N/A
Keywords: inter-organizational , statistical models , dynamics on networks , reciprocity
Accepted: Yes

275 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Dean Lusher
Authors: University of Melbourne Department of Psychology Parkville, VIC 3010, Australia
Garry Robins
University of Melbourne Department of Psychology Parkville, VIC 3010, Australia
Beliefs and networks: Personal attitudes and perceptions of general attitudes in exponential random graph (p*)
We define perceived beliefs as an individual’s perception of the attitudes of others within a network. Perceived
beliefs are people’s understandings of general attitudes held among their social partners and so may be an
indicator of norms within the network. Perceived beliefs can be measured as a node-level attribute and examined
with exponential random graph models for social networks. We explore the after-hours socialising relations of
players in four all-male professional football clubs. In three of the four clubs, the inclusion in models of both
perceived beliefs and personal attitudes towards masculinity demonstrates an alignment of sender and receiver
Abstract: effects. That is, individuals who perceive others as having more extreme masculine attitudes seek more social
partners; while individuals who personally have more extreme masculine attitudes tend to be more popular as
social partners. Such effects occur even after controlling for local structural effects, such as reciprocity,
transitivity, and general popularity and expansiveness processes. The implication of this sender-receiver
alignment is that individuals in the network take note of what they understand to be generally held attitudes, and
they seek to develop social links in line with these perceptions. The social network emerges in part from these
beliefs about what others are thinking.
Session: Exponential Random Graphs
Keywords: beliefs , attitudes , ergm (p*) , attributes
Accepted: Yes

276 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Susan Lynch
London Business School, Strategy and international management, 14 Oslo Court, Prince Albert Rd, London,
London Nw87EN, United Kingdom
Title: Organizational restructuring: formal boundaries, conflicting identities and tie
This paper considers how changes in formal structure and a key
element of informal structure -- the embeddedness of employee
relationships -- influence tie formation. When employees are moved
across formal groups, new relationships will form. Employees placed
in the same group form relationships based on new opportunities to
interact and to establish new organizational identities. Employees
separated into different groups form relationships based on common
Abstract: past experiences. Yet, moving employees also decreases the
structural alignment between the informal networks and the formal
position of moved employees. This conflict causes inconsistent
contributions to organizational identity, decreasing new tie
formation when employees are heavily embedded. I test my arguments
using a two-period panel of data on employee relationships in a
financial services provider. Results show evidence of an identity
conflict discount.
Session: Network Dynamics
dynamics on networks , egocentric networks , embeddedness , transitivity , intraorganizational networks ,
Accepted: Yes

277 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Jennifer MacGibbon
USMA, West Point, New York
Kelsey Tardieu
Authors: USMA, West Point, New York
Ian McCulloh, MS
USMA, West Point, New York
John Graham, PhD
USMA, West Point, New York
Title: Behavior Change due to Awareness of One’s Location in a Social Network
This is the first study in a five year strategic social network study that monitors all activity and transactions in a
group of twenty-five mid level management employees. The goal of this first study is to empirically evaluate
behavior change due to the awareness of one’s position in a social network and the difference between email
networks and one’s own perceived social network. This study hopes to expand on previous research of actual
and perceived networks (Krackhardt, 1990). Thus far, by monitoring anonymized email interaction, highly
sensitive data sets were collected and used for network analysis. Initial surveys of the employees provided data
for the perceived social network. It is hypothesized that participants who have few connections to other people in
the group are expected to change their behavior in order to make more connections, while those that have many
connections will not alter their behavior. Additionally, this study anticipates that email data network will strongly
resemble the actual network. Based on our primary finds, data thus far indicates that several participants play
formal roles within the network. We expect to find those same individuals playing vital positions in the email
network, as well as behavior shifts in the participants as social awareness increases.
Session: Actor-Network Theory and Science Studies
Keywords: affiliation networks , changes in networks , peer networks
Accepted: Yes

278 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Sofus Macskassy
Authors: sofmac@fetch.com
Fetch Technologies El Segundo, CA USA
Title: Graph Mining applied to Networks: Statistical Measures of Patterns and Clusters
This paper presents our recent work investing mining methods to help users understand large networks. Our
approach is based on statistical graph-measures for pattern analysis. Being able to quickly identify groups and
patterns in networks can significantly speed up the analysis and understanding of these. Specifically, we have
developed a novel statistical approach to pattern finding, which calculates meaningful similarity metrics that can
be used to rapidly identify common as well as unexpected patterns in networks. Our approach is flexible in that it
can identify a range of patterns from very abstract to very specific (e.g., isomorphisms).

We have evaluated our approach on a few publicly available networks from relational machine learning. This
Abstract: paper focuses on a relatively small network of hundreds of people, movies, production companies, etc. extracted
from the internet movie database (www.imdb.com). Specifically, we focused on patterns of size three (e.g., two
and a movie). Collapsing the network around instances of identified patterns (which we allow to overlap) results
in clusters which we can then analyze. Some of our interesting findings include the "actor country movie" pattern,
which collapsed into two large clusters, one consisting of U.S. and Europe and another with China and Hong
Kong. The "actor movie producer" pattern identified a large number of small clusters, suggesting that producers
generally keep to themselves or only co-produce with a limited set of other producers. This behavior interestingly
also appears to be the case with production companies.
Session: Algorithms and Analytic Methods
algorithm , data mining , graph theory , methodology , network structure , perceived social networks , social
network analysis tools
Accepted: Yes

279 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Balázs Mahler
Authors: bmahler@echosurvey.hu
Echo Survey Sociological Research Institute Székesfehérvár Forgó u. 15. 8000 Hungary
Title: The community-shaping role of illegal drugs
Echo Survey Sociological Research Institute made a survey research in spring 2006, on a high school aged
sample about their drug attitudes and using drugs. The research contained a multilevel network analysis to study
and redefine the community-shaping role of illegal drugs.
The positive impact of network analysis contrary to the traditional survey researches is the both qualitative and
quantitative characteristic, this method is between the micro and macro sociology. Our presentation’s goal is to
study how works the theory of using drugs and the relationships in a friendly community with network analysis
Professionals in drug issues say that using drugs has no effect on the quality (friendly) relations between young
adults. First, in our paper we analyze where drug addicts take place in the school class network and we try to
validate a significant difference between the drug users’ and the others’ personal relationships.
Second, we analyze the effects of personal drug attitudes on the friendly relationships with drug users, and study
the drug users role in their network, especially considering the prestige.
Session: social networks and health
Keywords: homophily , informal norms , social network , drug use
Accepted: Yes

280 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Aili Malm
Authors: amalm@csulb.edu
California State University, Long Beach, Criminal Justice, 1250 Bellflower Blvd, Long Beach, CA 90840, USA
Title: Criminal Network Habitats: Combining spatial and social network analysis in criminological research
This research introduces the concept of a criminal network habitat, the geographic space where criminal
networks are situated (through home, offense, and activity node location). The author examines the relationship
between social network and spatial variables of individuals (associates) involved in criminal networks in order to
explore the geographic and social space of criminal associates. The results show that the distance between
individuals in criminal networks and their associates vary systematically with network characteristics (centrality
measures) and crime type.
Session: Criminals, Gangs, Terrorists, and Networks
Keywords: centrality , georgraphic information system , criminal networks
Accepted: Yes

281 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Eric Mancik
USMA, West Point, NY, USA
Authors: Josh McCaskill
USMA, West Point, NY, USA
John Graham, PhD
USMA, West Point, NY, USA
Title: Laboratory Simulation of Hierarchical and Network Based Organizations
The focus of this first in a series of laboratory studies is to compare the effectiveness and efficiency of
hierarchical command and control (C2) vs. self-organizing, peer-based edge (E) organizational forms in
performing tasks that require decision making and collaboration. By using Experimental Laboratory for
Investigating Collaboration, Information-sharing, and Trust (ELICIT), an open-source software application based
on an intelligence analyst context and designed to test and observe interactions within different command and
control organizations and network models, we can evaluate the experimentally manipulated organization design
on performance. Further, an associated survey suite supports an understanding operations, information
processes and interactions between nodes within a network, allows a better understanding of the vital
interactions that establish and maintain a significant relationship.
Eighteen participants with varying levels of military experience, ranging from experts with possibly twenty years
or more to novices who have less than one served as role players in the two day study. The simulation goal is to;
given a dynamically distributed set of information about possible future terrorist attacks, identify the particulars of
the next attack. The simulation supports modeling of a standard military command and control organization as
well as a self-organizing network.
We found that organizational structure, ordering effects, and time in condition had impacts on organizational
performance and behavior. Further, we expect the unobstructed flow of information in the self-organizing network
will allow for different perspectives to analyze and identify potential adversary activities. By repetitively executing
these complex large organization studies over the next twelve months, we expect to find greater effectiveness
and participant comfort in the hierarchical structure. However, the self-organizing network results in greater
efficiency of information exchange.
Session: Simulation
Keywords: experiment , intelligence , terrorist , structure , information sharing
Accepted: Yes

282 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Maria Cristina Maneschy

Universidade Federal do Para (Federal University of Para State) - Departamento de Sociologia Belem - Para -
Klovdahl Alden
Australian National University School of Social Sciences Canberra - Australia
Conceicao Maria de Fatima
Universidade Federal do Para (Federal University of Para State) - Departamento de Sociologia Belem - Para -
Maia Maria Lucia Sa
Universidade Federal do Para (Federal University of Para State) - Departamento de Sociologia Belem - Para -
Title: Networks and social capital, the critical links: peasant associations’ in Eastern Amazon (Brazil)
In the northeastern Amazon region of Brazil there has been a notable proliferation of peasants’ associations,
particularly from the 1990s. Generally, they aim to provide alternative means to deal with common interests and
to organize individual and group efforts, thus contributing to reduced inequalities. At the same time, the deep
socioeconomic inequalities that these associations seek to reduce tend to limit their capacities to do so. Yet, the
magnitude of the ‘local associations’ phenomenon’, which is not limited to that region, has motivated many recent
studies and much debate.
From a social network perspective, it is plausible that the dynamics of the associations is influenced partially by
their contacts with different organizations, institutions and people outside the association. Here it is argued that
analyzing the social capital quality of social networks gains in efficacy from a twofold consideration of the
Abstract: networks structures and of the contents of the relationships from the participants’ points of view. So, together
with distances and pathways to contacts, analysis might also focus on communication processes, identifying
shared understandings and loyalties underneath interactions patterns.
This study draws on interviews with leaders of thirty-six rural associations in three municipalities in the
northeastern region of the State of Pará, Brazil. The field work was conducted in the first half of 2005 as part of
an ongoing research about the significance of local associations for promoting political participation for peasant
workers and their communities. The paper considers the extent to which these local organizations represent
forms of social capital for those involved, that is, sets of social relations or social networks able to produce and to
provide access to valued resources. The network environment of these local associations is examined, with
particular reference to their external links.
Session: Qualitative Social Networks Research
Keywords: network structure , qualitative network , social capital , amazonian region
Accepted: Yes

283 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Charles Mann
Department of Computer Science and Engineering Southern Methodist University P.O. Box 750122 Dallas, TX
75275-0122 USA
David Matula
Department of Computer Science and Engineering Southern Methodist University P.O. Box 750122 Dallas, TX
75275-0122 USA
Eli Olinick
Department of Engineering Management, Information and Systems Southern Methodist University P.O. Box
750123-0123 Dallas, TX 75275 USA
Title: Revealing the Hierarchical Community Structure of Social Networks Employing Concurrent Flow
A network flow algorithm based on fundamental principles of graph theory is introduced to identify the underlying
hierarchical community structure of a network. Finding a community structure that maximizes the Newman
modularity Q value is an NP-hard problem. We show that a reasonably-high Q value can be obtained in
polynomial time by partitioning the network in a succession of divisive cuts in the emerging largest component via
the maximum concurrent flow. Matula’s 1986 work establishing the Maximum Concurrent Flow Problem (MCFP)
and his papers on divisive vs. agglomerative average linkage hierarchical clustering provide the basis for
partitioning a social network by way of sparsest cuts and/or maximum concurrent flow.

The MCFP extends the Ford-Fulkerson source-sink max-flow min-cut theorem to all pairs maximum concurrent
flow. The density of an (S,T)-cut is | (S,T) | / ( | S | • | T | ) where |(S,T)| is the number of links between
communities S and T with | S | • | T | being the maximum number of links possible. The minimum density cut in
the network is the sparsest cut. Sparsest cuts of the largest components are iteratively determined via linear
programming until a multipartite cut is identified that is more constraining to concurrent flow than any sparse cut.
No extraneous links are identified.

Empirical results on random networks with embedded communities and selected social networks from the
literature are used to compare this concurrent flow algorithm to the Girvan-Newman algorithm that is based on
edge betweenness centrality.
Session: Algorithms and Analytic Methods
network structure , social networks , maximum concurrent flow , sparsest cut , divisive algorithm , graph
theory , multipartite cut , community structure
Accepted: Yes

284 of 500 2/16/2007 5:53 PM

INSNA Conference Papers https://cgi.sfu.ca/%7Einsna/confpapers/papersdetail.php

Drew B. Margolin
Authors: University of Southern California, Communication, 444 1/2 N. Curson Ave, Los Angeles, CA 90036, U.S.A.
Bettina M.R. Heiss
University of Southern California, Communication
Title: Ego Networks as Interactors: A New Level of Analysis in Network Evolution
Recent methodological advances in network science (Börner, Sanyal, & Vespignani, 2007) enable researchers to
study multilevel change in networks with increasing sophistication. Nevertheless, few theories provide a
systematic conceptual framework for investigating change as it occurs in these complicated cross-level
interactions, hence this progress in methodology is rarely harnessed sufficiently by network ana