Robustness of Latent Profile Analysis With Measurement Noninvariance Between Profiles

• In Event: Advanced Methods With Practical Implications

Abstract

Latent profile analysis (LPA) identifies heterogeneous subgroups based on continuous indicators that represent different dimensions. It is common practice to measure each dimension using items, create composite scores for each dimension, and use composite scores as indicators of profiles in LPA. In this case, measurement models for dimensions are not included and potential noninvariance across latent profiles is not modeled in LPA. This simulation study examined the robustness of LPA in terms of class enumeration and parameter recovery when the noninvariance was unmodeled by using composite scores as profile indicators. Results showed that factor mean difference, magnitude of noninvariance, and mixing proportions had impact on LPA. Implications will be provided for applied researchers on how to conduct LPA.

Authors

• Yan Wang, University of Massachusetts Lowell

• Eunsook Kim, University of South Florida

• ZHIYAO YI, University of South Florida