Measurement invariance in longitudinal social and behavioral research

May 28, 2015, 11:00 h
Mannheim, Conference room B2,8

Prof. Tenko Raykov


In this talk, a pragmatic definition of measurement invariance (MI) is initially discussed in the setting of longitudinal social and behavioral research. A strategy for searching for the roots of violations of MI when found in an empirical study is then outlined. A directly applicable latent variable modeling procedure is subsequently focused on, which allows one to test for MI and is based on an application of the increasingly popular Benjamini-Hochberg multiple testing method. The proposed procedure is illustrated with empirical data. The talk concludes with pointing out limitations of the procedure and some recommendations for longitudinal social and behavioral research concerned with invariance of studied latent constructs.

About the speaker

Tenko Raykov is Professor of Measurement and Quantitative Methods at Michigan State University at East Lansing, Michigan. In 1987, he earned his PhD in Mathematical Psychology at Humboldt University, Berlin, and completed his habilitation thesis in 1996. In his work, Tenko Raykov focuses on statistical modeling and psychometric theories with an emphasis on structural equation modeling.