Inferential Approaches for Network Analysis: AMEN for Latent Factor Models

Minhas, Shahryar; Hoff, Peter D.; Ward, Michael D.

doi:10.1017/pan.2018.50

Cited by 54 publications

(48 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the GSSRM and ERGM are not the only statistical approaches that model the structural dependencies of network data. On the contrary, numerous modelling alternatives are available that explicitly model the latent social structures that frequently typify social networks [ 90 , 91 ]. In the hands of behavioural ecologists, the use of these methodological tools can potentially prompt new theorizing and empirical tests about social behaviour.…”

Section: Discussionmentioning

confidence: 99%

Family ties: the multilevel effects of households and kinship on the networks of individuals

Koster¹

2018

R. Soc. open sci.

View full text Add to dashboard Cite

Among social mammals, humans uniquely organize themselves into communities of households that are centred around enduring, predominantly monogamous unions of men and women. As a consequence of this social organization, individuals maintain social relationships both within and across households, and potentially there is conflict among household members about which social ties to prioritize or de-emphasize. Extending the logic of structural balance theory, I predict that there will be considerable overlap in the social networks of individual household members, resulting in a pattern of group-level reciprocity. To test this prediction, I advance the Group-Structured Social Relations Model, a generalized linear mixed model that tests for group-level effects in the inter-household social networks of individuals. The empirical data stem from social support interviews conducted in a community of indigenous Nicaraguan horticulturalists, and model results show high group-level reciprocity among households. Although support networks are organized around kinship, covariates that test predictions of kin selection models do not receive strong support, potentially because most kin-directed altruism occurs within households, not between households. In addition, the models show that households with high genetic relatedness in part from children born to adulterous relationships are less likely to assist each other.

show abstract

Section: Discussionmentioning

confidence: 99%

Family ties: the multilevel effects of households and kinship on the networks of individuals

Koster¹

2018

R. Soc. open sci.

View full text Add to dashboard Cite

show abstract

“…The latent factor network approach infers this network by locating actors in a low‐dimensional “social space” (cf. Hoff 2005, 2015; Minhas, Hoff, and Ward 2019). The nature of the relationship between two actors is represented by a multiplicative effects term, which we call the relationship score.…”

Section: Measuring Cooperative and Conflictual Public Relationships Bmentioning

confidence: 99%

Elite Interactions and Voters’ Perceptions of Parties’ Policy Positions

Adams

Weschle

Wlezien

2020

American J Political Sci

View full text Add to dashboard Cite

Recent research documents that voters infer that governing coalition partners share similar ideologies, independently of these parties’ actual policy statements. We argue that citizens estimate party positions from more general forms of interparty cooperation and conflict, particularly near the times of national elections. We analyze tens of thousands of media reports on elite interactions from 13 Western democracies between 2001 and 2014, and show that—controlling for coalition arrangements and for the policy tones of parties’ election manifestos—voters infer greater left–right agreement between pairs of parties that have more cooperative public relationships, but that this “cooperation effect” is only evident near the times of national elections. Our findings have implications for parties’ policy images and for mass–elite linkages.

show abstract

“…This specification subsumes the models discussed previously in this section, by simply choosing A to be the identity matrix. Examples include Hoff (2008)'s eigenmodel for undirected networks, Hoff (2009)'s SVD-inspired extension to directed networks, and Minhas, Hoff and Ward (2016)'s class of additive and multiplicative effects (AME) models.…”

Section: Dot Product Modelsmentioning

confidence: 99%

The Geometry of Continuous Latent Space Models for Network Data

Smith¹,

Asta²,

Calder³

2019

Statist. Sci.

View full text Add to dashboard Cite

We review the class of continuous latent space (statistical) models for network data, paying particular attention to the role of the geometry of the latent space. In these models, the presence/absence of network dyadic ties are assumed to be conditionally independent given the dyads' unobserved positions in a latent space. In this way, these models provide a probabilistic framework for embedding network nodes in a continuous space equipped with a geometry that facilitates the description of dependence between random dyadic ties. Specifically, these models naturally capture homophilous tendencies and triadic clustering, among other common properties of observed networks. In addition to reviewing the literature on continuous latent space models from a geometric perspective, we highlight the important role the geometry of the latent space plays on properties of networks arising from these models via intuition and simulation. Finally, we discuss results from spectral graph theory that allow us to explore the role of the geometry of the latent space, independent of network size. We conclude with conjectures about how these results might be used to infer the appropriate latent space geometry from observed networks.

show abstract

Inferential Approaches for Network Analysis: AMEN for Latent Factor Models

Cited by 54 publications

References 33 publications

Family ties: the multilevel effects of households and kinship on the networks of individuals

Family ties: the multilevel effects of households and kinship on the networks of individuals

Elite Interactions and Voters’ Perceptions of Parties’ Policy Positions

The Geometry of Continuous Latent Space Models for Network Data

Contact Info

Product

Resources

About