Maximum likelihood estimation of a social relations structural equation model

Steffen Nestler,Alexander Robitzsch,Oliver Lüdtke

doi:10.1007/s11336-020-09728-z

Abstract

The social relations model (SRM) is widely used in psychology to investigate the components that underlie interpersonal perceptions, behaviors, and judgments. SRM researchers are often interested in investigating the multivariate relations between SRM effects. However, at present, it is not possible to investigate such relations without relying on a two-step approach that depends on potentially unreliable estimates of the true SRM effects. Here, we introduce a way to combine the SRM with the structural equation modeling (SEM) framework and show how the parameters of our combination can be estimated with a maximum likelihood (ML) approach. We illustrate the model with an example from personality psychology. We also investigate the statistical properties of the model in a small simulation study showing that our approach performs well in most simulation conditions. An R package (called srm) is available implementing the proposed methods.

Highlights

The social relations model (SRM) is widely used in psychology to investigate the components that underlie interpersonal perceptions, behaviors, and judgments
Results for the SR-confirmatory factor analysis (CFA) Table 1 shows the parameter estimates that we obtained with the two-step approach and the social relations model-confirmatory factor analysis (SR-CFA) for the two-factor model of the person-level effects
Our goal was to present a combination of the social relations model (SRM) with structural equation models (SEMs)

Summary

Introduction

The social relations model (SRM) is widely used in psychology to investigate the components that underlie interpersonal perceptions, behaviors, and judgments. The social relations model (SRM) is a conceptual and mathematical approach that can be applied to disentangle the components of interpersonal judgments and behaviors. It is used across different psychological disciplines (see Kenny et al 2006), including personality and social psychology, educational psychology, clinical psychology, and organizational psychology (see the SRM bibliography available at http://davidakenny.net/doc/srmbiblio.pdf). Limited because they were carved out to estimate a saturated model (i.e., all covariances between the round-robin variables are estimated) They cannot be applied to examine structural hypotheses concerning the multivariate relations between SRM effects measured by different round-robin variables. The researcher may be interested in examining whether the three liking variables are good measures of a latent perceiver-effect factor or a latent target-effect factor, respectively, or he or she may want to test whether a latent target effect for liking can be used to predict a latent target effect for extraversion

Objectives

Results

Conclusion