Abstract:

As the integration of programming and computational thinking (CT) into primary and secondary education becomes increasingly widespread, teachers are expected not only to acquire technical skills but also to develop positive attitudes toward programming as a subject and teaching domain. Many policy documents and professional development frameworks assume that teachers who are motivated, confident, and value programming are more likely to demonstrate stronger CT skills and, by extension, more effective teaching of these concepts. However, the empirical foundation supporting the alignment between affective factors and cognitive skill development in teachers remains limited. This study aims to investigate the extent to which in-service teachers’ programming attitudes are associated with their CT proficiency, using a multidimensional and data-driven approach.

Survey data were collected from 239 in-service teachers, including validated subscale scores for three programming attitude dimensions (Interest, Utility, Self-efficacy) and five CT skills (Abstraction, Algorithmization, Decomposition, Evaluation, and Generalization). Descriptive statistics and Pearson correlations were calculated, followed by Bayesian confirmatory factor analysis (CFA) to validate the measurement structure of both constructs. Canonical Correlation Analysis (CCA) was then used to examine multivariate relationships between the attitude and CT domains. For comparison and robustness, Bayesian multivariate regression models were also estimated.

Results demonstrated strong internal coherence within both constructs. CT dimensions showed moderate intercorrelations, consistent with the idea of a unified but multi-faceted skill set. Programming attitudes also demonstrated coherent internal structure, with Interest and Utility moderately associated and Self-efficacy forming a partially distinct attitudinal component. However, correlations between CT skills and programming attitudes were weak or negligible. Bayesian regression models confirmed that none of the attitude domains meaningfully predicted teachers’ CT performance. These findings suggest that positive orientations toward programming do not necessarily translate into higher CT competence among in-service teachers.

This study contributes novel evidence challenging the commonly assumed affective–cognitive alignment in teacher preparation for CT and programming. The results highlight the importance of treating technical skill development and attitude shaping as related but independent learning targets in professional development. Implications are discussed for the design of teacher education programs, curriculum planning, and assessment of teacher readiness in STEM and computing education.

Keywords:

Computational thinking, programming attitudes, teacher education, computing education.