Using the Graded Response Model to Validate and Standardize the Academic Self Concept Scale for University Students in Saudi Arabia

Abstract

This study aimed to standardize the Academic Self Concept (ASC) Scale for use with Saudi university students. The focus was on evaluating its psychometric properties, including reliability, validity, and ability to measure latent traits using Item Response Theory (IRT), specifically the Graded Response Model (GRM).Data were collected from 894 university students using the ASC scale, comprising 40 items across four domains: Mathematics, Verbal, Academic, and Problem-Solving. IRT-GRM was applied to assess item difficulty, discrimination, and test information. Structural Equation Modeling (SEM) was used to confirm the scale's construct validity, while measures marginal reliability was used to measure the scale’s reliability.IRT analysis revealed strong psychometric properties, with high item discrimination (0.40–1.68), good model fit indices, and high test reliability (marginal reliability = 0.93). The ASC scale demonstrated excellent psychometric properties. Construct validity was confirmed through SEM, SEM confirmed the four-factor structure, with strong fit indices (CFI = 0.96, TLI = 0.95, RMSEA = 0.07). Composite Reliability (CR), for the four domains were ranging from 0.85 to 0.92, served as evidence of convergent validity. Discriminant validity was achieved with values of HTMT < 0.85. The ASC scale demonstrated excellent psychometric qualities, supporting its use in measuring academic self-concept among Saudi students. Recommendations include its application in higher education for academic support programs and further validation in diverse populations. Future research should explore cultural influences and develop digital assessment tools.