Development and Validation of Mathematical Thinking Test Among Undergraduate Students in The Western Zone Kingdom of Saudi Arabia

Abstract

This study developed and validated a 50-item instrument to measure MT among undergraduate students in public universities in western KSA (Jeddah, Makkah, and Ta'if), addressing gaps in existing tools that rely on Classical Test Theory and target younger learners. Grounded in Item Response Theory (IRT) using the one-parameter logistic (Rasch) model, the instrument encompasses six dimensions: Generalization, Induction, Deduction, Use of Symbols, Logical Thinking, and Mathematical Proof. A cross-sectional survey design was employed with a stratified random sample of 229 participants. Content validity was established through expert review by six professors, yielding I-CVI scores of 0.833–1.00, S-CVI/Ave = 0.97, S-CVI/UA = 0.82, and excellent Fleiss' kappa agreement. Reliability analyses revealed Cronbach's alpha = 0.95, person reliability = 0.95 (separation = 4.25), and item reliability = 0.96 (separation = 5.14). Model-data fit confirmed unidimensionality (first-contrast eigenvalue = 1.8 < 3.0), local independence (residual correlations < 0.50), and productive item fit (infit/outfit MNSQ = 0.5–1.5). Differential item functioning (DIF) analysis identified minimal bias, with only two items flagged for gender and 4 for university location. The instrument aligns with Saudi Vision 2030 by providing a robust, psychometrically sound tool for enhancing pre-service teacher training, curriculum design, and MT assessment in higher education.