Assessing Students’ Computational Thinking AbilityBased on Mathematical Reasoning

Abstrak

Despite the growing emphasis on twenty-first-century competencies, assessment practices in mathematics education have not sufficiently captured theintegration of computational thinking and mathematical reasoning. This study aims to develop and validate the SMART-CT instrument as a multidimensional assessment framework for evaluating computational thinking within mathematical reasoning contexts among senior high school students. The instrument was developed using the ADDIE model and validated through expert judgment, second-order Confirmatory Factor Analysis (CFA), reliability analysis, and Multidimensional Item Response Theory (MIRT) based on data collected from 1,218 students. The results demonstrate that the instrument exhibits strong psychometric qualities, including satisfactory model fit, high internal consistency (α = 0.861), and adequate measurement precision within the ability range of θ = −0.5 to +1.0, although reduced sensitivity was observed at higher ability levels. By conceptualizing computational thinking as an integral component of mathematical reasoning and operationalizing it within a multidimensional measurement framework, this study contributes to the advancement of more authentic and process-oriented assessment practices in mathematics education.