Investigating Impact of Covariate Inclusion on Sample Size Requirements of Factor Mixture Modeling

• In Event: Experiment Designs of Educational Research Studies

Abstract

Factor mixture modeling (FMM) has been increasingly used in behavioral and social sciences to capture underlying population heterogeneity. However, a common issue that has been raised in applications of FMM is the sample size required for accurate class enumeration and parameter recovery. This Monte Carlo simulation study aims to investigate the issue of sample size requirements in FMM, particularly taking into account the impact of covariate inclusion on sample size requirements. Preliminary results indicated that sample size needed for accurate class enumeration depended upon effect size, magnitude of covariate effect, and mixing proportions. saBIC was recommended for model selection. Implications for applied researchers will be presented in the full paper.

Authors

• Yan Wang, University of Massachusetts Lowell

• Eunsook Kim, University of South Florida

• Hsien-Yuan Hsu, University of Massachusetts Lowell