Measurement Models in Social Science Research: A Data-Based Illustration of Four Confirmatory Factor Models and Their Conceptual Application

The available data, provided as a covariance matrix for easy integration into a structural equation modeling program, is example data to be used to replicate given examples provided in the following paper:<br><br>Confirmatory Factor Analysis (CFA) is a commonly used method to estimate latent factors among groups of observed variables and provides a valuable tool for social science researchers to examine factor validity. However, little attention has been given to the conceptual and theoretical implications of various CFA models. Evaluating multiple structural models is a careful balance between theoretical plausibility, parsimony, and quantitative tests of model fit. The current article uses data from a measure of empathy to illustrate the conceptual and theoretical application of four different CFA models: (1) a single-factor general empathy model, (2) a correlated four-factor model, (3) a higher-order model comprised of first-order factors and second-order general empathy latent factor, and (4) a bifactor model consistent of four first-order factors and a general empathy factor. Results from each model and their conceptual plausibility are presented. Syntax for all models in Mplus, R, Stata, and EQS programs are provided for reference. As familiarity with CFA and structural equation modeling methods grows, researchers must understand the theory-based implications of varying measurement models and test which best represent their data and explain their conceptual application.