Research utilizing the perspective of social networks can shed important light on political processes, as is illustrated by the other articles in this issue and also the special issue Social Networks and American Politics of American Politics Research (Heaney and McClurg, 2009). This perspective offers some complications for statistical analysis, however. A network approach is so useful because it can represent the interdependence between political actors (see Huckfeldt, 2009) – but statistical modeling is commonly based on independence assumptions. The challenge in statistical modeling of social network data is to represent the dependencies between network ties so that valid inferences can be obtained and misspecification avoided; and, by doing so, to provide methods that allow researchers to test hypotheses about these interdependencies. This article treats statistical methods for network panel data. It is assumed that the reader has a basic knowledge of networks and the associated terminology; see, e.g., Wasserman and Faust (1994) or Knoke and Yang (2008). For the data structure it is assumed that a fixed set of nodes is being considered – where, however, exceptions are allowed in the sense that some nodes may enter or leave the network – while the change

Most existing research about online trust assumes static trust relations between users. As we are informed by social sciences, trust evolves as humans interact. Little work exists studying trust evolution in an online world. Researching online trust evolution faces unique challenges because more often than not, available data is from passive observation. In this paper, we leverage social science theories to develop a methodology that enables the study of online trust evolution. In particular, we propose a framework of evolution trust, eTrust, which exploits the dynamics of user preferences in the context of online product review. We present technical details about modeling trust evolution, and perform experiments to show how the exploitation of trust evolution can help improve the performance of online applications such as rating and trust prediction.

There is a growing interest in intra- and interorganizational network dynamics. The central question in the latter domain is, ‘‘How do firms choose collaborative partners given their pre-sent network configuration, their goals, and characteristics, to get a strategic network posi-tion?’ ’ We introduce actor-oriented network models as a method to describe and explain the development of interorganizational collaboration networks. The models are applied to longitu-dinal data about collaborative agreements within the genomics industry.

In this paper, we present a collective classification approach for identifying untrustworthy individuals in multiagent communities from a combination of observable features and network connections. Under the assumption that data are organized as independent and identically distributed (i.i.d.) samples, traditional classification is typically performed on each object independently, without considering the underlying network connecting the instances. In collective classification, a set of relational features, based on the connections between instances, is used to augment the feature vector used in classification. This approach can perform particularly well when the underlying data exhibits homophily, a propensity for similar items to be connected. We suggest that in many cases human communities exhibit homophily in trust levels since shared attitudes toward trust can facilitate the formation and maintenance of bonds, in the same way that other types of shared beliefs and value systems do. Hence, knowledge of an agent’s connections provides a valuable cue that can assist in the identification of untrustworthy individuals who are misrepresenting themselves by modifying their observable information. This paper presents results that demonstrate that our proposed trust evaluation method is robust in cases where a large percentage of the individuals present misleading information.

III Stochastic models for network dynamics 6 Basic model definition....................... 7

Abstract. In this paper, we present a collective classification approach for identifying untrustworthy individuals in multi-agent communities from a combination of observable features and network connections. Under the assumption that data are organized as independent and identically distributed (i.i.d.) samples, traditional classification is typically performed on each object independently, without considering the underlying network connecting the instances. In collective classification, a set of relational features, based on the connections between instances, is used to augment the feature vector used in classification. This approach can perform particularly well when the underlying data exhibits homophily, a propensity for similar items to be connected. We suggest that in many cases human communities exhibit homophily in trust levels since shared attitudes toward trust can facilitate the formation and maintenance of bonds, in the same way that other types of shared beliefs and value systems do. Hence, knowledge of an agent’s connections provides a valuable cue that can assist in the identification of untrustworthy individuals who are misrepresenting themselves by modifying their observable information. This paper presents results that demonstrate that our proposed trust evaluation method is robust in cases where a large percentage of the individuals present misleading information.

Abstract: The development and spread of research methods in sociology can be understood as a story about the increasing sophistication of tools in order to better answer fundamental disciplinary questions. In this chapter we argue that recent developments, related to both increased computing power and data collection ability along with broader cultural shifts emphasizing interdependencies, have positioned Social Network Analysis (SNA) as a powerful tool for empirically studying the dynamic and processual view of schooling that is at the heart of educational theory. More specifically, we explore how SNA can help us both better understand as well as reconceptualize two central topics in the sociology of education: classroom interaction and status attainment. We conclude with a brief discussion about possible future directions network analysis may take in educational research, positing that it will become an increasingly valuable research approach because our ability to collect streaming behavioral and transactional data is growing rapidly. 1

am deeply grateful to three people who helped me in my graduate studies and made this work possible. Foremost, my advisor David Knoke was a source of inspiration and guidance. He provided enough structure to keep me focused yet enough independence to develop my research skills and interests. His expertise in organizations and networks, statistics and research guidance were what made completion possible. Most of all, he was an ideal mentor and advocate on the West Bank, and I cannot imagine having gone through this with anyone else. Next, I thank Father Dennis Zehren for granting me permission to do research in his parish and for his wonderful example of trusting in God. His devout faith inspired me to develop the aspect of trust in God in this dissertation. My request for access to parish information, staff and families came at the worst time in the total stewardship program, yet he made a leap of faith, opened the doors and made this research possible. My husband John Reandeau is a constant source of encouragement, love, trust and support. He is the most critical one in helping me reach and obtain my goals and inspires me to better myself. Thanks for attempting to keep the house together while I was finishing. While the results were mixed, I appreciate the sentiments. I also want to acknowledge some key people in my graduate studies. I thank my

Author: Fenell, Minna Title: Communication as a trust builder in change management

Copyright c © by Societa ̀ editrice il Mulino, Bologna. Tutti i diritti sono riservati.