AN INVESTIGATION INTO THE MEASUREMENT INVARIANCE OF PISA 2012 MATHEMATICAL LITERACY TEST

The aim of this study is to investigate the measurement invariance of PISA 2012 mathematical literacy among the OECD member countries, and gender and region groups in Turkey. Among cognitive test booklets implemented in PISA 2012, booklet 8 which was used commonly by all countries was selected for this correlational survey study. The study was conducted using the dataset belonging to 23.311 students that took booklet 8 in the OECD member countries and 377 students that took this booklet in Turkey. Initially, measurement models were verified for all groups by performing a confirmatory factor analysis separately for OECD member countries, gender and region groups. The research then proceeded with the phase of testing the measurement invariance by testing the equivalence of covariance matrices for all groups. The measurement invariance was tested by means of a multi-group confirmatory factor analysis. The results revealed that the measurement invariance held true for the gender and region groups in Turkey, but not for the OECD member countries, and that the strong factorial invariance model was the one that worked most successfully among the models of measurement invariance