Identification and Scaling of Latent Variables in Ordinal Factor Analysis

TL;DR

This paper introduces minimal 'integer constraints' for ordinal factor analysis, aligning latent variable predictions with the average of ordinal variables, and demonstrates their effectiveness through proofs, data illustrations, and simulations.

Contribution

It proposes a novel minimal identification constraint called 'integer constraints' that simplifies ordinal factor analysis by aligning latent variables with observed ordinal data.

Findings

Integer constraints lead to intuitive parameterizations.

Integer constraints perform similarly to other methods in simulations.

The approach is validated with real data examples.

Abstract

Social science researchers are generally accustomed to treating ordinal variables as though they are continuous. In this paper, we consider how identification constraints in ordinal factor analysis can mimic the treatment of ordinal variables as continuous. We describe model constraints that lead to latent variable predictions equaling the average of ordinal variables. This result leads us to propose minimal identification constraints, which we call "integer constraints," that center the latent variables around the scale of the observed, integer-coded ordinal variables. The integer constraints lead to intuitive model parameterizations because researchers are already accustomed to thinking about ordinal variables as though they are continuous. We provide a proof that our proposed integer constraints are indeed minimal identification constraints, as well as an illustration of how integer constraints work with real data. We also provide simulation results indicating that integer constraints are similar to other identification constraints in terms of estimation convergence and admissibility.