<![CDATA[This study examines the structure and predictors of student performance in designing unplugged coding student worksheets integrated with environmental issues in a coding instruction course. This study employed a peer assessment approach to examine the interrelationships among Cognitive Engagement (CE), Behavioural Engagement (BE), Affective Engagement (AE), Academic Responsibility (AR), Collaboration (COL), and Communication (COM), drawing upon theories of student engagement, cooperative learning, computational thinking, peer assessment, and self-regulated learning. A total of 100 pre-service elementary school teachers enrolled in the course participated as respondents. Data were collected using a 12-item Likert-scale peer assessment rubric and analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) via SmartPLS4. Results of the outer model confirmed convergent validity (outer loadings ≥ 0.70; AVE ≥ 0.50) and composite reliability (CR ≥ 0.843). The inner model revealed that Affective Engagement significantly predicted Academic Responsibility (β = 0.606, T = 3.845, p < 0.001) with a large effect size (f² = 0.375), and Behavioral Engagement significantly predicted Collaboration (β = 0.571, T = 3.407, p = 0.001) with the largest effect size (f² = 0.598). The model explained 55.2%, 72.7%, and 67.3% of the variance in Academic Responsibility, Collaboration, and Communication, respectively, with a GoF index of 0.78, indicating a strong overall model fit. These findings highlight the pivotal role of affective and behavioural dimensions of engagement in shaping collaborative competence and academic responsibility among pre-service teachers engaged in computational thinking design tasks.]]>
