Session Summary

Examining Measurement Invariance of Background Questionnaire Indicators in International Large-Scale Assessment

Mon, April 16, 4:05 to 5:35pm, New York Hilton Midtown, Floor: Third Floor, Americas Hall 1-2 - Exhibit Hall

Session Type: Poster Session

Abstract

International large-scale assessment programs have developed detailed background questionnaires under an explicit mandate to measure key constructs that are believed to explain student academic performance. These programs face the challenge of obtaining comparable measures across diverse national and cultural contexts, which are characterized by differences from many factors such as language, culture, educational system, and experience of examinees. This paper will investigate the relationship between measurement invariance and the prediction accuracy of background indicators in explaining science performance in PISA 2015. We will allow different degrees of cross-national measurement invariance. The main focus will be the effect of the measurement invariance assumption on cross-national comparisons of the fixed effects and random effects of the generalized linear mixed model.

Sub Unit

Papers