Testing Measurement Invariance in Multilevel Data With Unequal Within-Level and Between-Level Factor Structures

• In Event: Poster Session 14
  In Poster Session: Advanced Issues in Multilevel Factor Analysis

Abstract

This simulation study is focused on testing multilevel measurement invariance at the cluster level using the design-based and multiple-group multilevel confirmatory factor analysis (MMCFA) approaches. Three factor structures are examined: (1) the same within- and between-level structure, (2) complex within and simple between structure, and (3) simple within and complex between structure. Results are evaluated based on inadmissible solution rates, power, and type 1 error rates under various simulation conditions, including factor structures, number of clusters, cluster size, intraclass correlation (ICC), and size of non-invariance. Preliminary results show that both approaches show good power of detecting noninvariance when the ICC is relatively large. With a small ICC, the power is quite low especially for the design-based approach.

Authors

• Lihua Yang, University of Arkansas at Fayetteville

• Xinya Liang, University of Arkansas

• Wen-Juo Lo, University of Arkansas