Investigation of measurement invariance of Science Motivation and Self-Efficacy Model: PISA 2015 Turkey sample

Measurement invariance analyses are carried out in order to find evidence for the structural validity of the measurement tools using in the field of educational sciences and psychology. The purpose of this research is to examine the measurement invariance of Science Motivation and Self-Efficacy Model constructed by Instrumental Motivation to Learn Science and Science Self-Efficacy subscales in the PISA 2015 Student Questionnaire across different groups in Turkey sample. The study was carried out by means of 4583 students' data which meet the assumptions of the analyses. The measurement invariance of the model in terms of gender and statistical region groups was examined with structural equation modeling (SEM) technique. Firstly, data was examined whether the necessary assumptions for the analyses were met. Then, measurement models were verified by performing confirmatory factor analysis (CFA). The measurement invariance across genders and statistical regions was tested by multi-group confirmatory factor analysis (MGCFA). Unweighted Least Squares (ULS) method was used as the estimation method in CFA and MGCFA stages. The results of the study show that the research model ensures only configural invariance across gender groups and weak invariance across regions. Science Motivation and Self-Efficacy Model illustrates that it is not valid to make comparisons between scores of male and female students or students from different regions of Turkey. According to findings, the research model could not provide complete measurement invariance. Researchers are advised to be cautious when comparing scores in affective domain in science by different groups.