PLS-SEM Approach: Validity and Reliability of a Questionnaire on Context, Input, Process and Acceptance of STEM Implementation in Malaysia

Abstract

This study aims to test the validity and reliability of the context, input, process, and acceptance of STEM implementation questionnaires for secondary school teachers in Malaysia. Partial Least Squares Structural Equation Modelling (PLS-SEM) with SMARTPLS 3.2.3 software was used to do the analysis. The study involved 825 secondary school teachers using stratified random sampling. Based on composite reliability and Cronbach's alpha values, the Reflective Measurement Model showed that there was internal consistency reliability. Convergent validity was also achieved through outer loading analysis and average variance extracted (AVE). It also met the criteria for discriminant validity based on the Fornell-Larcker Criterion, Cross Loading, and the Heterotrait-Monotrait (HTMT) Ratio. The findings of this study demonstrate that this questionnaire is valid and reliable for assessing STEM implementation among secondary school teachers in Malaysia.