Fueled by recent advances in statistical modeling and the rapid growth of network data, social network analysis has become increasingly popular in sociology and related disciplines. However, a significant amount of work in the field has been descriptive and correlational, which prevents the findings from being more rigorously translated into practices and policies. This article provides a review of the popular models and methods for causal network analysis, with a focus on causal inference threats (such as measurement error, missing data, network endogeneity, contextual confounding, simultaneity, and collinearity) and potential solutions (such as instrumental variables, specialized experiments, and leveraging longitudinal data). It covers major models and methods for both network formation and network effects and for both sociocentric networks and egocentric networks. Lastly, this review also discusses future directions for causal network analysis. Expected final online publication date for the Annual Review of Sociology, Volume 48 is July 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

The recent surge in inequality has been linked to growing disparities both within and between economic strata. Existing approaches to analyzing changes in inequality, however, often ignore within-group inequality by solely analyzing mean differences between groups. Approaches that do allow examining both changes in within- and between-group inequality in turn are limited in addressing causal questions about why inequality is changing. This paper introduces a novel approach to analyzing how a treatment variable affects both changes in within- and between-group inequality and decomposing these changes into compositional and behavioral effects. The procedure combines a classic variance decomposition with the Kitagawa-Blinder-Oaxaca (KBO) decomposition approach. Compared to KBO, however, the method allows analyzing treatment effects not only on the mean but on the whole conditional distribution. I demonstrate the utility of the approach with an application analyzing the changing impact of motherhood on women’s earnings and its consequences for earnings inequality.

In this paper, I propose a new statistical method to represent the complex multilevel structure of coalition government data. The method opens up opportunities to model and test theories on the interplay of agents in their collective impact on entities, which bears the potential to advance our understanding of other research areas characterized by interdependence structures, such as multi-party wars, treaties, and international organizations. I model the interdependence structure by endogenizing the weights of a multiple membership multilevel model (MMMM). This approach allows for a flexible weighting of coalition parties in their effect on governments conditional on the interdependencies among them. To the best of my knowledge, this is the first time that the weights of an MMMM have been endogenized. With this paper, I provide the R package ‘Rmm’ to estimate such models for a variety of outcomes (linear, logit, conditional logit, Cox, Weibull) in a user-friendly way.

The causal identification of network effects is fraught with theoretical and statistical difficulties. To guide social network researchers in choosing the right empirical strategy, this survey of network effect models (i) maps out the different types of network effects and levels of analysis, (ii) discusses data and measurement issues related to network effect identification, and (iii) reviews the key challenges and solutions to causally identifying network effects. The survey covers both modeling solutions based on random assignment and network models for observational data.

As for students many consequential life decisions still lie ahead it is vitally important that their choices suit their abilities. Concerning education a misperception of academic ability can lead to educational misinvestment with potentially severe consequences. That is why this paper investigates if there are disparities in the ability to accurately self-evaluate school performance by social origin. To the best of the author’s knowledge, this is the first paper considering this important research question. In doing so, the paper has two emphases: firstly, a theoretical model, arguing why disparities in the ability to accurately self-evaluate school performance by social origin are likely, is proposed and secondly an empirical study is conducted in order to examine if disparities by social origin are findable. The key results indicate that both students with less and students with highly educated parents underestimate their school performance if they have school grades higher than the average, and overestimate their school performance if they have school grades lower than the average. However, this relationship is intensified for students with less educated parents and therefore they self-evaluate their school performancecompared to students with highly educated parents less accurately.