The study of social network dynamics has become an increasingly important component of many disciplines in the social sciences. In the past decade, statistical models and methods have been proposed which permit researchers to draw statistical inference on these dynamics. This thesis builds on one such family of models, the stochastic actor oriented model (SAOM) proposed by Snijders [2001]. Goodness of fit for SAOMs is an area that is only just beginning to be filled in with appropriate methods. This thesis proposes a Mahalanobis distance based, Monte Carlo goodness of fit test that can depend on arbitrary features of the observed network data and covariates. As remediating poor fit can be a difficult process, a modified model distance (MMD) estimator is devised that can help researchers to choose among a set of model elaborations. In practice, panel data is typically used to draw SAOM-based inference. This thesis also proposes a score-type test for time heterogeneity between the waves in the panel that is computationally cheap and fits into a convenient, forward model selecting workflow. Next, this thesis proposes a rigorous method for aggregating so-called relational event data (e.g. emails and phone calls) by extending the SAOM family to a family of hidden Markov models that suppose a latent social network is driving the observed relational events. Finally, this thesis proposes a measurement model for SAOMs inspired by error-in-variables (EiV) models employed in an array of disciplines. Like the relational event aggregation model, the measurement model is a hidden Markov model extension to the SAOM family. These models allow the researcher to specify the form of the mesurement error and buffer against potential attenuating biases and other problems that can arise if the errors are ignored.

The structure of social networks, and people's position within these networks, are important predictors of many individual and group-level outcomes. One type of social networks that is regularly studied are the mutually interdependent relations of friends. This thesis focusses on friendship networks between adolescents in the context of schools. Arguably the most important and consistently found regularities in adolescent friendship networks are i) the tendency of friendships to be mutual, called reciprocity; ii) their tendency to cluster in groups, known as transitivity; and iii) the tendency of friendships to be present between those that are similar to one another, called homophily. Various social theories originating in different disciplines have theoretically proposed and empirically found micro-mechanisms that explain the regular occurrence of these substructures in friendship networks. This thesis introduces a framework of how the relation between these different networks tendencies can be understood. I propose that each of the three network evolution mechanisms can be connected to a type of social situation in which friends interact to form and maintain their friendships. Social situations that are dyadic and only involve two persons are connected to reciprocal friendships. Group-based social situations, on the other hand, are related to transitivity and homophily, where the groups are either defined socially or through common characteristics. Starting from this proposition, I suggest that when two adolescents share one forum for interaction with one another, i.e. they regularly meet within one of the social situations, meeting in additional other situations does not increase the likelihood of a friendship tie existing as much as could be expected from the sum of the effect of meeting in either situation. Consequently, I expect a negative interaction between the different network mechanisms. After a series of empirical analyses that support the outlined reasoning, I use the developed perspective to investigate how the micro-mechanisms contribute differentially to the creation of newly formed friendships and to the maintenance of already existing friendships. Finally, I show how a situational understanding of friendship can be used to differentiate which friendships are most important for social influence and for peer pressure.

Social network analysis (SNA) provides tools to examine relationships between people. Text mining (TM) allows capturing the text they produce in Web 2.0 applications, for example, however it neglects their social structure. This paper applies an approach to combine the two methods named "content-based SNA" (CB-SNA). Using the R mailing lists, R-help and R-devel, we show how this combination can be used to describe people's interests and to find out if authors who have similar interests actually communicate. We find that the expected positive relationship between sharing interests and communicating gets stronger as the centrality scores of authors in the communication networks increase. / Series: Research Report Series / Department of Statistics and Mathematics

How accurately people perceive interpersonal relationships, both among others and with themselves, forms the basis of social inferences about the structure of the social environment and one's place in it. Six hypotheses were tested using the cognitive social structures method from social network analysis with five independent but similar student networks from two universities. Results from all networks were meta-analyzed. Participants gave both their self-reported friendship ratings for every alter in their group and also gave their perceptions of the ratings the other member would give. Perception ratings were correlated to self-report ratings for each participant as a measure of accuracy of social network perception. Participants perceived more structural balance than was present in self-reports in four out of five networks and in the meta-analysis, providing evidence for the balance schema. Attachment anxiety correlated negatively with accuracy for one of the networks but was not statistically significant in the meta-analysis. Being located in a tightly-knit subgroup reduced overall network accuracy, consistent with the strength of weak ties (SWT) theory, in one network but not in the meta-analysis. In only one network did participants overestimated how central they were, though not significantly in the meta-analysis. Being more central in the social network was unrelated to accuracy, as was the mean social network distance between perceiver and targets. Results provide meta-analytic support for the balance schema and limited support for attachment, SWT, and egocentric bias in social network perception.

Fisheries straddling or migrating between international maritime boundaries represent a typical case of the tragedy of the commons. Over two dozen Regional Fisheries Management Organisations (RFMOs) have been created to manage these fisheries, which means it also represents a typical case of 'regime complexity' or 'governance architecture'. These literatures recognise that such institutions do not operate independently and therefore institutional functions such as attracting participants, practising their regulatory role, and performing their mandate should be understood as interdependent. This thesis proposes that we study such institutions together with actors and architectures of relations between and among them, which together I term 'governance complexes', by means of a relational approach. This relational approach combines relational theory, which posits the operation of endogenous relational mechanisms alongside exogenous explanations such as institutional design, with network methods that enable structural insights and robust inference that takes into account these interdependencies. The dissertation comprises two main parts that describe and explain the global fisheries governance complex, respectively. The first describes how the governance complex's three main components, states, RFMOs, and states participation in these RFMOs, have evolved. A topological typology utilising key network concepts is proposed and employed to show that the global fisheries governance complex is not fragmenting but becoming more overlapped and nested. The second part explains how this governance complex has evolved in terms of (1) participation, (2) practice, and (3) performance. First, it finds that while states find institutional design features such as an RFMO’s internal organisation attractive, relational mechanisms such as popularity and closure also provide important heuristics for participatory decisions in complex settings. Second, it finds that high levels of organisation also enables higher regulatory activity, but so do relational mechanisms such as coercion or imitation. Third, it proposes the concept of net effectiveness to gauge an institution’s "take-home" performance once its position in the broader governance complex has been taken into account. The result is not only an explanation of the evolution of global fisheries governance but also a developmental step towards an institutional relationalist theory of governance complexes.

Doctor of Philosophy / Security Studies / David Graff / How does the social network structure of China’s Central Committee influence the direction and timing of intra-Party events, domestic policies, and foreign policies? How do changes in network structures explain specific patterns and propensities for policy change? The purpose of this study is to describe the social network structure of the Chinese Communist Party’s Central Committee from 1922 to 2011 and to explore the relationship between changes in the network and policy trends. This study draws on policy network theory, network dynamics, Veto Player Theory and Prospect Theory which together posit that dynamic network structural characteristics influence policy outcomes. I introduce a dynamical theory of policy networks and describe some observable implications. This mixed method analysis is based on a new network dataset and follows two major lines of investigation. A structured, focused comparison of case studies associates changes in the Central Committee’s network structure in 1969, 1978, 1982, and 1997 with consequent policy outcomes and demonstrates the explanatory power of the dynamical theory. Statistical analyses of China’s foreign policies (1949-1978 and 1984-1995) and China’s domestic policies (1984-1995) suggest the dynamical theory is generalizable. Changes in paramount leader and potential competitor centralities and Central Committee centralization are important causal factors explaining the timing and type of intra-Party, domestic, and foreign policies.

A presente tese apresenta, discute e avalia criticamente o conteúdo e o contributo de uma seleção de trabalhos de pesquisa submetidos ou publicados em revistas internacionais no campo das ciências do esporte. Coletivamente, esses trabalhos contribuem para o campo da análise de jogo em jogadores jovens e adultos de futebol, com implicações para aperfeiçoar a avaliação e o treinamento. As pesquisas foram em parte moldadas pela experiência do autor na prática profissional e acadêmica, mas principalmente impulsionada por necessidades emergentes identificadas durante os anos de atuação prática e em parceria com profissionais atuantes em clubes de futebol. Um total de duas revisões sistemáticas (apresentadas na presente tese como introdução ao campo de estudo), sendo uma metaanálise, e três estudos de campo revisados por pares estão incluídos. Esta tese apresenta criticamente resultados e discussões sobre dois principais fluxos: (i) validação de testes físicos em ambiente de jogo, de acordo com as características gerais e exigências específicas das partidas; (ii) análise dos fatores contextuais e posicionais que afetam potencialmente o desempenho no futebol. Os trabalhos de pesquisa são apresentados em uma sequência conceitual dentro dos dois temas, em vez de uma ordem estritamente cronológica, para demonstrar a coerência e a sinergia dentro dos dois fluxos. Esta tese fornece uma reflexão crítica sobre a contribuição geral para o corpo atual de conhecimento científico e o impacto coletivo dos trabalhos que foi alcançado. As limitações dos estudos encontradas no decorrer do trabalho são discutidas como temas atuais, possibilitando um direcionamento para futuras pesquisas. Por fim, todo conhecimento produzido foi traduzido em forma de um e-book para maior acesso por diferentes recursos humanos / This thesis presents, discusses and critically evaluates the content and the contribution of a selection of research papers submitted or published to international peer-reviewed sports science journals. Collectively, these papers make novel insights to the field of match analysis in young and adults\' soccer players, with implications for improving assessment and training. The researches were partly shaped by the authors\' experience in professional and academic practice, but mainly driven by emerging and evolving needs-analyses identified within his work, identified during the years of practice and in partnership with professionals working in soccer teams. A total of two systematic reviews (presented in this document as an introduction to the field of study), one of them being a metaanalysis, and three original articles are included. These papers present critically comments on work in two main streams: (i) validation of physical fitness tests in a game context, according to general characteristics and specific requirements of the match; (ii) analysis of contextual and positional factors that potentially affect the soccer performance. The original research papers are presented in a conceptual sequence within the two themes, rather than a strictly chronological order to demonstrate coherence and synergy within the two collections. The document provides critical reflection on the general contribution to the current body of scientific knowledge and the collective impact of the work that has been achieved. The limitations of the studies found during the course of the work are discussed as current themes, allowing a direction for future research. Finally, all scientific knowledge produced was translated into an e-book for greater access by different human recourses

In 2011, Oregon was one of many states in the U.S. consolidating their education governance around an early learning, K-12, and postsecondary hub. This study uses legislator-voting data to investigate the relationship between this consolidated model and endogenous policy formulation processes. This study employs a separable temporal exponential random graph model (STERGM) to investigate how an education governance shift toward consolidated authority relates to bipartisan outcomes for education-related bills over time. Oregon legislator voting networks were analyzed for cohesion, centrality, and community detection measures, as well as by legislator attributes (e.g. gender, party, and title) to test the association they had on the likelihood of forming ties with other legislators. Finally, to study the relationship of bipartisanship with legislators’ likelihood to vote commonly, I added the legislators’ political party attributes within dyads to analyze the association that having different political parties had on legislators’ common votes. The results highlight evidence of legislator networks that were very dense at each time point included in the study, with a high likelihood of forming ties. However, when Oregon shifted to centralized education governance model their legislator networks became more distributed and cohesive when compared to other years included in the longitudinal study. It is possible that such a shift prompted collaboration among legislators resulting in mutuality that increased the likelihood for underrepresented groups of legislators (e.g. females and republicans) to vote commonly with their colleagues. Aligned with previous research, this study found that centralized governing bodies reinforced by political legislation provided collaborative initiatives for the legislative community. Attending to bipartisan voting patterns dynamically through a governance shift is a valuable investigation that will provide nuanced inferences about education governance and policymaking for states making similar consolidated governance shifts in the future.

Propomos o uso da análise de redes sociais para entender efeitos de agenda-setting. Para tanto, recuperamos a evolução dos estudos sobre a teoria da agenda-setting, desde as aplicações relacionadas à mídia tradicional até os estudos contemporâneos sobre mídias sociais. Em seguida, apresentamos o conceito de agenda-setting interpessoal e descrevemos métodos da análise de redes sociais para investigação desse conceito. A operacionalização empírica da proposta foi feita pela comparação de dois tipos de redes: 1) um grupo de indivíduos que publicou mensagens no Twitter sobre um tópico transmitido na TV nacional e 2) um grupo de indivíduos que publicou mensagens no Twitter sobre um tema que não fez parte da mídia broadcast. Nosso objetivo foi investigar as diferenças entre as redes de indivíduos que experimentaram efeitos de agenda-setting advindos da mídia de massa e indivíduos que experimentaram efeitos de agenda-setting interpessoal. Nossos resultados ainda são preliminares, mas indicam diferenças estruturais entre os grupos e evidências de que indivíduos mais centrais são impactados antecipadamente por mensagens transmitidas interpessoalmente. / We propose the use of social network analysis to understand agenda-setting effects. To achive that purpose, we recover the evolution of studies on agenda-setting theory, from applications related to traditional media to contemporary studies on social media. Next, we present the concept of interpersonal agenda-setting and describe methods of social network analysis to investigate this concept. The empirical operationalization of the proposal was made by comparing two types of networks: 1) a group of individuals that tweeted about a topic broadcasted on national TV and 2) a group of individuals that tweeted about a non-mainstream topic. We aim to investigate differences between networks of individuals that experienced mass media agenda-setting effects and individuals that experienced interpersonal agenda-setting effects. Our results are still preliminary but indicate structural differences between the groups and also indicate evidence that more central individuals are impacted in advance by interpersonally transmitted messages.

Thesis advisor: Larry Ludlow / In this dissertation study, I examined the extent to which a relational network of teachers, administrators, two-way immersion (TWI) experts and mentors promote productive communities of practice (CoP). In a conventional instruction, teachers are often isolated in their classrooms, and a private practice culture prevails. In 2012, the Two-Way Immersion Network for Catholic Schools (TWIN-CS) was launched in an effort to support school reform by engaging school leaders and teachers to collectively learn toward implementing TWI models in their schools. Using the framework of communities of practice (Lave & Wenger, 1991; Wenger, 1999), I employed a case study design (Yin, 2009) to explore a national network of Catholic elementary school educators. Data sources included qualitative data featuring semi-structured interviews and quantitative source from a relational network survey. Qualitative results revealed that organizational features of TWIN-CS are critical in promoting participants’ learning to implement TWI. In particular, participants discussed the annual TWIN Summer Academy and bi-monthly webinars to be instrumental for their learning. Many participants also shared that an expansion of CoPs beyond TWIN-CS further prompted productive learning. However, the qualitative evidence also showed a lack of clear internal and external network structures and role definition, and sustaining connection beyond the Summer Academy and webinars were perceived as a great challenge. Quantitative results suggest that TWIN-CS has a core-and-periphery network structure with the Boston College design team at the innermost core, with visibly dense ties connecting to and from them. Most teachers, on the other hand, occupy the most peripheral positions in this network. Survey evidence also showed that participants generally perceived a much stronger learning relationship within schools and showed less certainty on cross-network relationships. In terms of learning characteristics, majority of the respondents viewed knowledge sharing, trust, and advice-oriented dimensions “strongly” but perceived a lack of data-driven learning for both within school and cross-network. I conclude this study with a discussion of implications for future research and practice. / Thesis (PhD) — Boston College, 2016. / Submitted to: Boston College. Lynch School of Education. / Discipline: Educational Research, Measurement and Evaluation.