Unpacking the Psychometric Properties of Computational Thinking Measure (Stage 2, 1:59 PM)

• In Event: e-Lightning Ed-Talk Session 18 - Stage 2

Abstract

This study examined the latent structure of an 11-item computational thinking (CT) assessment administered to middle school students. Guided by information processing theory and multidimensional models of CT, we used exploratory and confirmatory factor analysis to investigate whether student responses reflected a unidimensional, two-factor, or bifactor structure. Results supported a two-factor solution, with items loading onto distinct but moderately related dimensions. A bifactor model showed acceptable fit but did not significantly improve model performance. Findings suggest that CT, as operationalized in this instrument, is best represented as a multidimensional construct. Implications include the value of domain-specific scoring and instruction that target diverse CT subskills in middle-grade STEM education.

Authors

• Shaf Ahmed, Utah State University

• David F. Feldon, Utah State University