ABSTRACT

Purpose This study evaluates program learning outcomes in a Physical Education Teacher (PET) program by integrating validated assessment data with cluster-based classification. Methodology Using the CIPPIEST program evaluation framework, the product/outcome component was examined through graduation examination results and compulsory course performance data from 67 students who completed the PET program (2020–2024) at the Mongolian National University of Education. Classical test theory was applied to analyze the difficulty, discriminant power, and reliability of randomly selected test items from the graduation examination. Results The results indicated acceptable psychometric properties (p = 0.79–0.81; D = 0.54–0.57; r_pbi = 0.93–0.96), supporting the reliability of the assessment data. Based on integrated knowledge and skills performance across general, teacher education, professional, and practicum courses, a two-step cluster analysis classified graduates into five distinct learning attitudes. The findings show that 80.7% of graduates were classified within higher learning attitudes clusters, while 19.3% were grouped within comparatively lower learning attitudes. Implications for research and practice The study demonstrates that integrated performance patterns in knowledge and skills can serve as an empirical basis for identifying learning attitudes in program evaluation contexts. The results provide evidence-based implications for curriculum revision and targeted professional development within PET.