Estimating Social Influence Using Latent Space Adjusted Approach in R

TL;DR

This paper demonstrates how a latent space adjusted approach in R can effectively disentangle social influence effects from confounding factors like homophily, improving estimation accuracy in social network analysis.

Contribution

It introduces a practical implementation of the latent space adjusted method for estimating social influence, addressing estimation biases in social network models.

Findings

The approach reduces bias in social influence estimation.

Simulation results show improved accuracy over existing methods.

Empirical example illustrates practical application in R.

Abstract

Social influence, sometimes referred to as spillover or contagion, have been extensively studied in various empirical social network research. However, there are various estimation challenges in identifying social influence effects, as they are often entangled with other factors, such as homophily in the selection process, the individual's preference for the same social settings, etc. Methods currently available either do not solve these problems or require strong assumptions. Recent works by Xu 2018 and others show that a latent-space adjusted approach based on the latent space model has potential to disentangle the influence from other processes, and the simulation evidence shows the approach performs better than other state-of-the-art approaches in terms of recovering the true social influence effect when there is an unobserved trait co-determining influence and selection. In this paper we illustrate how latent space adjusted approach accounts for bias in the estimation of the social influence effect, and demonstrate how this approach can be implemented to estimate various social influence models with an empirical example in R.