Evaluating the impact of policy interventions on respondents who are embedded in a social network is often challenging due to the presence of network interference within the treatment groups, as well as between treatment and non-treatment groups. In this paper, we propose a novel empirical strategy that combines network sampling based on the identification of independent sets with a stochastic actor-oriented model (SAOM) to infer the direct and net effects of a policy. By assigning respondents from an independent set to the treatment, we are able to block direct spillover of the treatment among the treated respondents for an extended period of time, during which the direct effect of the treatment can be isolated from the associated network interference. We empirically demonstrate this using a simulation-based evaluation of a fictitious policy implementation using both real-life and generated networks, and use a counterfactual approach to estimate the treatment effect of the policy. Our results highlight the effectiveness of our proposed empirical strategy, and notably, the role of network sampling techniques in influencing the evaluation of policy effects. The findings from this study have the potential to help researchers and policymakers with planning, designing, and anticipating policy responses in a networked society.

This pre-registered meta-analysis uses multi-level random effects models to give precise estimates of average outbound best friend and friend nominations – average outdegree centrality – in youth friendship networks and examines whether average outdegree centrality varies by age, gender, and geographic region. Pooling 196 estimates reported in 51 articles reflecting 37 datasets from whole network studies, youth nominated 4.80 best friends on average (SE=.37). Additionally, pooling 64 estimates reported in 20 articles reflecting 18 datasets from whole network studies, youth nominated 6.05 friends on average (SE=.60). Early adolescents (10–14 years) nominated significantly more best friends than adolescents (15–18 years). However, there were no significant differences in average outdegree centrality by the gender or geographic region of the sample. Findings provide future research directions for understanding youth socializing environments and implications for peer interventions.

Statistical analysis of bipartite networks frequently requires randomly sampling from the set of all bipartite networks with the same degree sequence as an observed network. Trade algorithms offer an efficient way to generate samples of bipartite networks by incrementally ‘trading’ the positions of some of their edges. However, it is difficult to know how many such trades are required to ensure that the sample is random. I propose a stopping rule that focuses on the distance between sampled networks and the observed network, and stops performing trades when this distribution stabilizes. Analyses demonstrate that, for over 650 different degree sequences, using this stopping rule ensures a random sample with a high probability, and that it is practical for use in empirical applications.

The present empirical study aims to explore medical knowledge sharing in the Australian healthcare context, aiming to broadly evaluate the potential impact of Project ECHO®, an online mentoring and networking health program. We focus on health-related knowledge sharing practices among the network of professionals through formal and informal channels, and across different health and non-health sectors and organisational systems. Studying knowledge transmission among professional networks is essential for optimizing healthcare delivery, promoting innovation, and providing insights on improvement of patient experiences within the healthcare system. We utilize a multilevel approach to shape our data collection strategy. Employing network measures and Multilevel Exponential Random Graph Models, we aim to explore how advice and knowledge sharing behaviours among healthcare professionals and their institutions are interdependently connected. Then, we incorporate network generated results within an evaluation framework for establishing some aspects of the efficiency of the ECHO program along four pillars: Acceptability, Capability, Reachability, and Integration. Our investigation found that among ECHO members, hierarchy is less pronounced compared to across levels and organizations, with certain individuals emerging as central in advice-sharing. The multilevel network perspective showed complex, informal patterns of knowledge and information sharing, including inter-organizational hierarchy, role and sector homophily, brokerage roles with popularity across health organizations, and connectivity through knowledge-sharing in cross-level small group clusters.

BackgroundWe examined the role of network communication about HIV-related topics in mediating the efficacy of a social network intervention on HIV seroconversion among people who inject drugs (PWID) in Ukraine, where Eastern Europe’s second-largest HIV epidemic is concentrated among PWID. MethodsWe used randomized controlled trial data from 1200 HIV-negative PWID (Ukraine; 2010–2012) in an inverse-odds weighted analysis to examine mediation by network communication. ResultsNetwork communication mediated 24 % (95 % CI= 19.22–29.38) of the intervention’s effect. ConclusionsIntegrating training to support network communication about additional HIV prevention resources could enhance the impact of social network HIV prevention interventions among PWID.

Social norms influence homophobic behavior, yet these norms are often misperceived. We study the extent to which friendship ties and group memberships are related to misperceptions of opinions towards homosexuality, and how these misperceptions are sustained in social networks through opinion sharing. We find that misperceptions lead individuals to be less willing to share their opinions with ethno-religious ingroup members, non-friends or with individuals whom they perceive to hold different opinions. Although differences observed in the context of this study are relatively small, they may add up over time. These results offer scope for interventions that try to reduce norm misperceptions between groups - as a way to stimulate social change towards a more tolerant society.

Surveys conducted on social groups often generate incomplete information due to imperfect response rates. Drawing on Facebook data from a nationally representative sample of graduating college students in Taiwan, we examined the extent to which partial contact records predict which Facebook users belong to a specific class. We first used data from classes with low to middle response rates to train a model for classmate prediction. Based on data from classes with high or perfect response rates, we simulated data by using four different sampling methods with various response rates, and applied the trained model on simulated data to classmate prediction. With a minimal response rate of 40 percent, we achieved an accuracy rate of 90 percent and a true positive rate of 86 percent. Chronological order sampling had the best prediction performance, followed closely by popularity sampling, then by random sampling, and lastly by unpopularity sampling.

We suggest a novel approach to determining the centrality measures for directed signed networks, based on the notion of social balance. We postulate that along with the existing positive connections, the structure of positive and negative connections can be used to determine potential secondary connections, respectively, weak social ties between pairs of individuals who are, e.g., either friends with the same person or under threat from the same person. This kind of connection agrees perfectly with the theory of social balance. Given the structure of primary and secondary connections, the centrality is measured using an eigenvector-based scheme. The suggested approach is applied to the classical example of the social network of monks in a monastery, and the results show a good agreement with the available ground truth.