Regression-based Negative Control of Homophily in Dyadic Peer Effect Analysis

TL;DR

This paper introduces a regression-based negative control method to accurately estimate peer effects in social networks, addressing homophily bias that confounds causal inference in contagion studies.

Contribution

It develops a novel regression approach using negative controls to identify and estimate contagion effects despite homophily bias, advancing causal analysis in social network data.

Findings

Effective in isolating contagion effects from homophily bias

Applied to Framingham data to assess peer influence on obesity

Provides a robust method for causal inference in social network studies

Abstract

A prominent threat to causal inference about peer effects over social networks is the presence of homophily bias, that is, social influence between friends and families is entangled with common characteristics or underlying similarities that form close connections. Analysis of social network data has suggested that certain health conditions such as obesity and psychological states including happiness and loneliness can spread over a network. However, such analyses of peer effects or contagion effects have come under criticism because homophily bias may compromise the causal statement. We develop a regression-based approach which leverages a negative control exposure for identification and estimation of contagion effects on additive or multiplicative scales, in the presence of homophily bias. We apply our methods to evaluate the peer effect of obesity in Framingham Offspring Study.