Measurement Invariance of Self-Regulated Learning Perception Scale in Medical Students.

Abstract

BACKGROUND Measurement invariance analysis is a critical step in evaluating the structural validity of measurement instruments, ensuring that a scale measures its intended constructs equally across different subgroups. Such analyses are essential before making subgroup comparisons. This study investigated the measurement invariance of the Self-Regulated Learning Perception Scale (SRLPS), a tool designed to assess self-regulated learning levels among medical students. The SRLPS consists of 4 categories: motivation and action to learning, planning and goal setting, strategies for learning and assessment, and lack of self-directedness. MATERIAL AND METHODS A cross-sectional study was conducted involving 878 medical students spanning the first to final years. Multigroup confirmatory factor analysis (MG-CFA) was employed to assess measurement invariance across genders and years of medical education, categorized as preclinical and clinical years. Invariance was evaluated by comparing model fit indices across successive stages of invariance testing. RESULTS The findings demonstrated that the SRLPS achieved full measurement invariance in both sexes, confirming its suitability for valid comparisons between male and female medical students, with ΔCFI ≤0.01. For year of medical education, configural invariance was established, indicating that the 4-factor structure of the SRLPS is consistent across preclinical and clinical students. However, only partial metric invariance was achieved, suggesting that factor loadings for some items differ between preclinical and clinical groups. CONCLUSIONS The measurement invariance should be evaluated before using a scale for subgroup comparison. The SRLPS can be used to make valid comparisons between male and female medical students, but cannot be used for comparison between preclinical and clinical year students.