Causal inference in collaboration networks using propensity score methods

Abstract Using panel data of school-class networks of 11–13-year-old students, this study investigates effects of schoolwork collaboration-networks on grades and school-related well-being. It suggests propensity score weighting-regression as a method of causal inference for data collected in social contexts, and in studies analyzing node-attributes as outcomes of interest. It will argued that this alternative approach is useful when stochastic actor-based models (SAOMs) show convergence problems in sparse networks. Three methods of causal analysis dealing with the problems of endogeneity bias and interference between observations will be discussed in this study: first, SAOMs for the co-evolution of networks and behavior/attitudes will be estimated, but this results in a systematic loss of data. Second, propensity score matching compares treated cases with untreated nearest neighbors. However, the stable-unit-treatment-value assumption (SUTVA) requires that the analysis controls for network embeddedness in the final analysis. This is possible by using propensity score weighting-regression, which is a flexible approach to capture treatment diffusion via multiplex networks.