Volume 28, Issue 1, 2008

Cover and Publication Information, 1
Table of Contents, 2-3
Rethinking Preferential Attachment Scheme: Degree centrality versus closeness centrality, 4-15
Ko,Kilkon, , Lee,Kyoung,Jun , Park,Chisung,
Construction of realistic dynamic complex networks has become increasingly important. One of the more widely known approaches, Barabasi and Albert’s “scale-free” network (BA network), has been generated under the assumption that new actors make ties with high degree actors. Unfortunately, degree, as a preferential attachment scheme, is limited to a local property of network structure, which social network theory has pointed out for a long time. In order to complement this shortcoming of degree preferential attachment, this paper not only introduces closeness preferential attachment, but also compares the relationships between the degree and closeness centrality in three different types of networks: random network, degree preferential attachment network, and closeness preferential attachment network. We show that a high degree is not a necessary condition for an actor to have high closeness. Degree preferential attachment network and sparse random network have relatively small correlation between degree and closeness centrality. Also, the simulation of closeness preferential attachment network suggests that individuals’ efforts to increase their own closeness will lead to inefficiency in the whole network.
How Correlated Are Network Centrality Measures?, 16-26
Valente,Thomas,W , Coronges,Kathryn, , Lakon,Cynthia, , Costenbader,Elizabeth,
Analysis of Transitivity and Reciprocity in Online Distance Learning Networks, 27-39
Aviv,Reuven, , Erlich,Zippy, , Ravid,Gilad,
The goal of this research is to explore whether network structures common in social networks also emerge in online distance learning networks. We compare the observed values of reciprocity and transitivity in 95 online distance learning networks and 40 social networks with predictions of Random Graph models. All the networks tested exhibit reciprocities that significantly deviate from the predictions of the random graph models. All the online distance learning networks and some, but not all, of the social networks exhibit transitivity compatible with generalized random graph models. We provide possible explanations for these behaviors, based on the broadcast nature of the online distance learning networks, and practical implications for online collaborative learning.
The Life Cycle of Collaborative Partnerships: evolution of structure and roles in industry-university research networks., 40-58
Trotter,Robert,T , Briody,Elizabeth,K , Sengir,Gülcin,H , Meerwarth,Tracy,L
Global corporations have initiated collaborative partnerships with university research institutions, private-sector firms, and other strategic partners at an increasingly rapid pace over the last decade. These partnerships create collaborative networks that leverage knowledge acquisition and technology transfer necessary to keep corporations and universities at the cutting edge of competition. Consequently, corporations have a competitive need to be able to predict the ideal structure, dynamics, and life cycles of these partnerships in order to effectively initiate, maintain, repair, and exit them in a way that retains the potential for future collaboration for both sides of the partnership. This paper provides an empirically validated model of the evolutionary structures and role relationships found in successful collaborative partnerships. The research combined ethnographic methods with qualitative and quantitative social network paradigms to identify the key structural frameworks and role configurations critical to the health of partnerships over their typical life cycle. The results include a description of the structures and the key player dynamics of these partnerships through six life cycle stages (approach, initiation, start-up, growth, maturity, and transition).
The Eurovision Song Contest as a ‘Friendship’ Network, 59-72
Dekker,Anthony,
This paper examines the votes cast in the 2005 Eurovision Song Contest. Adjusting votes for song quality, a friendship network with valued links is obtained. Statistical analysis shows that friendship between countries is largely determined by geographical proximity, with a visible five-bloc structure. However, large immigrant groups often swayed national ties by voting for their home country. Some countries, such as Switzerland, appear to play a significant bridging role, and the Eastern Mediterranean bloc appears to act as a bridge to the new Balkan countries. Analysis thus reveals an emerging Europe very different from previous network studies of this kind. The analysis techniques demonstrated here have more general applicability, and may be useful for analysing other types of friendship networks.
Where Does Help Come From: A Case Study of Network Analysis in an Academic Group, 73-87
Li,Pengxiang, , Xi,Youmin, , Yao,Xiaotao,
This paper explores the relationship between the transferring activities of such resources as information, knowledge, social support, and tie strength of interaction among individuals in an academic group. We investigated the academic group led by Professor Xi in the School of Management of Xi’an Jiaotong University through questionnaire. The empirical data for four kinds of networks such as help, communication, friend, and research cooperation were collected and analyzed. Findings show that the interaction ties in the group cannot be viewed simply as strong ties, but the complex situation with strong and weak ties interweaved. Among the activities of transferring resources, nearly sixty percent of help came from strong ties and about sixty-five percent of help providers were the core actors in the group. This means that strong ties are the main channels of resource transfer while the role of weak ties should still be given attention. The strong ties with the core actors in the group can make it easier to get more valuable help than weak ties. However, new information and possible opportunities are more likely to be provided through weak ties than strong ties.

Volume 28, Issue 2, 2008

td>
Cover, 1-3
Looking at Social Capital through Triad Structures, 4-16
Prell,Christina, , Skvoretz,John,
The concept of social capital, which has gained wide currency in the literature, examines how actors’ ties to others advantage or disadvantage them and the groups to which they belong. Two conceptually distinct types of social capital, closure and brokerage in Burt’s (2005) terms, have been identified. In this paper, we propose a method by which brokerage and closure can be distinguished using a census of patterns of ties in triads of actors. We apply this method to network data gathered on 24 non-profit organizational actors. Our findings show when a network is characterized by brokerage or closure and how that network coincides with the presence/absence of trust and reciprocity. We conclude with a discussion on the nature of non-profits, and how the larger social context of network actors, in this case non-profits, play a role in interpreting the network structures uncovered via social network analysis.
Change in Connectivity in a Social Network over Time: A Bayesian Perspective, 17-27
Adams,Susan, , Carter,Nathan, , Hadlock,Charles,
In this paper, we propose a Bayesian methodology for examining differences between statistics of a social network at two distinct points in time. The problem has been of interest for some time in the social networks community because it is quite difficult to test whether differences over time in statistics such as overall network connectivities are significant. Several issues make this problem challenging: links in a social network tend to be dependent, and the networks at the two different points in time are likely to be dependent as well. This implies, for example, that bootstrapping a social network to address this problem may be impractical. This paper expands on a previously published Bayesian version of the p1 model for social networks with random effects, which allows for dependence between the edges of the networks. We use the software Winbugs to obtain posterior distributions for the difference in connectivity over time and for the correlation between each actor’s connectivities in the network at both points in time. We assume that this correlation is the same for all actors. We illustrate our methods with the case of a social network of collaborations (joint publications) between departments of a business university where interdisciplinary work was actively promoted. Our methods allow us to compare the tendency to make collaborative links across departments before and after the administrative initiatives.
Determining An Actor’s Network Capacity And Network Utilization: A Markov Model Of Human Agency In Social Networks, 28-42
Sundararajan,Binod,
The focus of this paper has been to put forth a Markov model that will provide information to actors in a network about the optimum capacity of alters in their social networks and how to maintain social-temporal relations with their contacts and resources with optimum efficiency. This model takes advantage of the similarities between the concept of human agency and Markov random processes. It takes into account the fact that present experiences are the sum total of past iterational and habitual experiences and the present practical-evaluative capacity to evaluate these past experiences. The model then adapts the Markov process and the Erlang blocking formula used in telephony, when it uses the present practical-evaluative experience to create a projective capacity toward the future state of the social network. This will then provide the actor (ego) with information about the efficient utilization of his/her channel/network capacity and the number of contacts or resources he/she would need to maintain to achieve the future stability of the network contacts and resources.
Scan Statistics for Interstate Alliance Graphs, 43-64
Marchette,David,J , Priebe,Carey,E
This paper discusses work on graphs defined in terms of alliances between countries. Scan statistics are used to investigate years in which there are an unusual number of agreements, not just between one country and its allies, but amongst the allies themselves. This is related to work on email ``chatter'' discussed in Priebe et al. (2005). The scan statistic detects unusually high (or low) values for a graph invariant within a local region of the graph (an induced subgraph). Thus, without a priori knowledge of where in the graph the detection might occur, we seek to detect a region of the graph that is very different from the other regions. We will use a particular graph invariant, the size, or number of edges in the graph, to help detect interesting changes in the alliance graphs that we investigate. We will be more precise below, but the idea is as follows: A detection at scale 0 corresponds to a single country making an unusually large number of alliances; a detection at scale 1 corresponds to a country and its allies making a large number of alliances among themselves. This can be a measure of the cohesiveness of the group; a detection at scale 2 (and higher) corresponds to a larger spreading of the alliances. It means that not only are there more alliances among the countries allied with the central country, but among their allies there are more alliances. This paper seeks to perform two tasks: the first is to introduce scan statistics to those in the social network community not familiar with this work; the second is to determine whether, in the case of interstate alliances, there are any interesting detections at scales above 0. We will demonstrate that sometimes this type of behavior is interesting.
Preference or Propinquity? The Relative Contribution of Selection and Opportunity to Friendship Homophily in College, 65-80
Godley,Jenny,
This paper examines the relative importance of preference and propinquity as determinants of socio-demographic homophily in friendship choice among students at a small college in the Northeastern United States. Using unique retrospective data, the paper first assesses friendship homophily over the four years of college. Friendship is homophilous across gender and race. QAP regression is used to determine the impact of both preference and propinquity (measured by participation in joint extra-curricular activities and shared academic major) on friendship choice. While preference predicts friendship choice in the freshman year, propinquity remains the strongest determinant of friendship choice over the four years.
Consent and Confidentiality: Exploring Ethical Issues in Public Health Social Network Research, 81-96
Harris,Jenine,K
Current ethical regulations were necessarily developed in response to unethical treatment of human subjects by clinical and social researchers in settings ranging from Nazi concentration camps in the 1940s to U.S. government offices in the 1960s. Due to a focus on relationships, social network studies pose complex ethical dilemmas regarding consent and confidentiality that often challenge these ethical regulations. These issues have kept social network projects from receiving Institutional Review Board (IRB) approval, and, in the case of Virginia Commonwealth University, halted human subjects research university-wide. In public health, social network analysis is an effective method for understanding how diseases are transmitted, how health messages are spread, how social support impacts morbidity and mortality, and how public health organizations collaborate. A review of 50 public health articles using social network approaches showed that few authors discussed issues of consent and confidentiality. Without accessible examples of how others have addressed consent and confidentiality, these issues will continue to challenge public health social network researchers and their IRBs.

Connect with Us

Facebook Twitter

Our Mission

INSNA is the professional association for researchers interested in social network analysis. The association is a non-profit organization incorporated in the state of Delaware and founded in 1977.

Contact:

support@insna.org

+1 (304) 208-8001 Tel
+1 (304) 523-9701 Fax

c/o JulNet Solutions, LLC
      1404 1/2 Adams Avenue
      Huntington, WV 25704