Examination of Measurement Invariance at Groups' Comparisons: A Study on PISA Student Questionnaire


Basusta N. B. , GELBAL S.

HACETTEPE UNIVERSITESI EGITIM FAKULTESI DERGISI-HACETTEPE UNIVERSITY JOURNAL OF EDUCATION, vol.30, no.4, pp.80-90, 2015 (Peer-Reviewed Journal) identifier identifier

  • Publication Type: Article / Article
  • Volume: 30 Issue: 4
  • Publication Date: 2015
  • Journal Name: HACETTEPE UNIVERSITESI EGITIM FAKULTESI DERGISI-HACETTEPE UNIVERSITY JOURNAL OF EDUCATION
  • Journal Indexes: Social Sciences Citation Index, Scopus, TR DİZİN (ULAKBİM)
  • Page Numbers: pp.80-90

Abstract

In order to make a meaningful comparision, instruments must have similar measurement qualities across settings. Research on measurement invariance (MI) has typically focused on internal structure of multi-item instruments and external relations. The purpose of this study is to investigate the factors of PISA student questionnaire's items and assessing the measurement invariance of these factors across gender. This study examined the PISA data for Turkish students with the sample size of 4942 through the analysis of Structural Equation Modeling (SEM). First, it is designed a measurement model over the factors from student questionnaire with Structural equation modeling. It was understood that the designed measurement model's goodness fit index output is good. Then four forms of invariance were assessed for each latent variables progressively and assessed MI. The variables were examined using comparative fit index (CFI). The MI results revealed that there were no invariance problems across gender in PISA.