Multiple imputation for missing edge data: A predictive evaluation method with application to Add Health

Summary1. Animals are embedded in dynamically changing networks of relationships with conspecifics. These dynamic networks are fundamental aspects of their environment, creating selection on behaviours and other traits. However, most social network-based approaches in ecology are constrained to considering networks as static, despite several calls for such analyses to become more dynamic.2. There are a number of statistical analyses developed in the social sciences that are increasingly being applied to animal networks, of which stochastic actor-oriented models (SAOMs) are a principal example. SAOMs are a class of individual-based models designed to model transitions in networks between discrete time points, as influenced by network structure and covariates. It is not clear, however, how useful such techniques are to ecologists, and whether they are suited to animal social networks. 3. We review the recent applications of SAOMs to animal networks, outlining findings and assessing the strengths and weaknesses of SAOMs when applied to animal rather than human networks. We go on to highlight the types of ecological and evolutionary processes that SAOMs can be used to study. 4. SAOMs can include effects and covariates for individuals, dyads and populations, which can be constant or variable. This allows for the examination of a wide range of questions of interest to ecologists. However, high-resolution data are required, meaning SAOMs will not be useable in all study systems. It remains unclear how robust SAOMs are to missing data and uncertainty around social relationships. 5. Ultimately, we encourage the careful application of SAOMs in appropriate systems, with dynamic network analyses likely to prove highly informative. Researchers can then extend the basic method to tackle a range of existing questions in ecology and explore novel lines of questioning.

show abstract

“…the ‘Held‐Out Predictive Evaluation’ of Wang et al . ). Methods that generate uncertainty around observations of animal networks (e.g.…”

Section: Challenges When Using Saoms To Study Animal Networkmentioning

confidence: 97%

Analysing animal social network dynamics: the potential of stochastic actor‐oriented models

Fisher

Ilany

Silk

et al. 2017

Journal of Animal Ecology

View full text Add to dashboard Cite

show abstract

“…Nevertheless, ERGMs may be biased when complete network data are not present (e.g., Robins, Pattison, & Woolcock, 2004). As discussed in detail in the Supporting Information, we considered modeling approaches (e.g., Robins et al, 2004) or imputation (Wang, Butts, Hipp, Jose, & Lakon, 2016) to deal with missing data. However, these methods rely on the presence of in-degree nominations from other actors, and we did not have consent to use their data in this way.…”

Section: Analytic Strategymentioning

confidence: 99%

A Social Network Analysis of Friendship and Spiritual Support in a Religious Congregation

Todd

Blevins

2019

American J of Comm Psychol

View full text Add to dashboard Cite

Religious congregations are social settings where people gather together in community to pursue the sacred (Pargament, 2008). Such settings are important to understand as they provide a context for individuals to develop relationships, share ideas and resources, and connect individuals to larger society (Todd, 2017a). Yet, research to date has not deeply examined the inherently relational nature of religious congregations. Thus, in this study, we used social settings theory (Seidman, 2012; Tseng & Seidman, 2007) to develop and test hypotheses about relationships within one Christian religious congregation. In particular, we used social network analysis to test hypotheses about relational activity, popularity, and homophily for friendship and spiritual support types of relational links. Our findings demonstrate how relational patterns may be linked to participation in congregational activities, occupying a leadership role, a sense of community and spiritual satisfaction, stratification, socialization, and spiritual support. Overall, this advances theory and research on the relational aspects of religious congregations, and more broadly to the literature on social settings. Limitations, directions for future research, and implications for theory and religious congregations also are discussed.

show abstract

“…Observed existent ties were coded 1, observed nonexistent ties were coded 0, and unobserved ties were coded as missing ("NA" in R). This approach to imputation is derived from the approach proposed by Handcock and Gile [42,48], and has been used to impute unobserved ties in other studies [52,53].…”

Section: Fitting a Model For Imputation Of Unobserved Friendshipsmentioning

confidence: 99%

Using Partially-Observed Facebook Networks to Develop a Peer-Based HIV Prevention Intervention: Case Study (Preprint)

Khanna¹,

Goodreau²,

Michaels³

et al. 2018

Preprint

View full text Add to dashboard Cite

Background: This is a case study from an HIV prevention project among young Black men who have sex with men (YBMSM). Individual-level prevention interventions have had limited success among YBMSM, a population that is disproportionately affected by HIV; peer networkbased interventions are a promising alternative. Facebook is an attractive digital platform because it provides the ability to broadly characterize social networks. There are, however, several challenges in using Facebook data for peer interventions, including the large size of Facebook networks, difficulty in assessing appropriate methods to identify candidate PCAs, boundary specification issues, and partial observation of social network data.Objective: We aim to explore methodological challenges in using social Facebook networks to design peer network-based interventions for HIV prevention, and present techniques to overcome these challenges.Methods: Our sample contained 298 "uConnect" study respondents who answered a biobehavioral survey in person and whose Facebook friend lists were downloaded (2013)(2014). The study participants had over 180,000 total Facebook friends who were not involved in the study ("nonrespondents"); we did not observe friendships between these nonrespondents. Given the large number of nonrespondents whose networks were partially observed, a relational boundary was specified to select nonrespondents who were "well connected" to the study respondents and who may be more likely to influence the health behaviors of YBMSM. A stochastic model-based imputation technique, derived from the exponential random graph models (ERGMs), was applied to simulate 100 networks where unobserved friendships between nonrespondents were imputed. To identify PCAs, the eigenvector centrality and keyplayer positive algorithms were used; both algorithms are well-suited to identifying individuals in key network positions for information diffusion. For both algorithms, we assessed: the sensitivity of identified PCAs to the imputation model; the stability of identified PCAs across the imputed networks, and the effect of the boundary specification on the identification of PCAs.Results: All respondents and 79% of nonrespondents selected as PCAs by eigenvector on the imputed networks were also selected as PCAs on the observed networks. For keyplayer, the agreement was much lower: 43% and 33% of respondent and nonrespondent PCAs respectively selected on the imputed networks were also selected on the raw network. Eigenvector also produced a stable set of PCAs across the 100 imputed networks and was much less sensitive to the specified relational boundary. Conclusion:While we do not have a gold standard that tells us which algorithm produces the most optimal set of PCAs, the lower sensitivity of eigenvector centrality to key assumptions leads us to conclude that it may be preferable. The methods we employed to address the challenges in using Facebook networks may prove timely given the rapidly increasing interest in using online social networks to impr...

show abstract

Multiple imputation for missing edge data: A predictive evaluation method with application to Add Health

Cited by 60 publications

References 23 publications

Analysing animal social network dynamics: the potential of stochastic actor‐oriented models

Analysing animal social network dynamics: the potential of stochastic actor‐oriented models

A Social Network Analysis of Friendship and Spiritual Support in a Religious Congregation

Using Partially-Observed Facebook Networks to Develop a Peer-Based HIV Prevention Intervention: Case Study (Preprint)

Contact Info

Product

Resources

About