Understanding Multiple Pathways to Computational Thinking: Emotional Trajectories and Interaction Patterns in Pre-Service Teacher Education

Pre-service teachers entering computer education programs often possess adequate computational skills yet struggle to transform this technical knowledge into effective classroom pedagogy. This exploratory mixed-methods study examined how cognitive, affective, and behavioral processes co-evolve during computational thinking (CT) development in a Thai pre-service teacher education context, addressing three research questions: (1) what emotional trajectories characterize CT development during an 8-week flipped classroom intervention; (2) how do self-efficacy, cognitive load, and anxiety interact to shape CT outcomes; and (3) what HCI patterns distinguish successful from struggling learners. We tracked 30 second-year Computer Education pre-service teachers at a regional Thai university, collecting 240 weekly psychological self-reports, 18 phenomenological interviews, continuous LMS analytics, 180 screen-recorded problem-solving sessions, and social network analysis of 847 forum posts. Three emotional trajectories emerged: Progressive Adapters (43%) reduced anxiety linearly through peer support; Fluctuating Persisters (37%) maintained engagement despite emotional volatility; and Struggling Stabilizers (20%) converted persistent anxiety into intensive help-seeking. All three groups achieved comparable CT gains (95.9%), suggesting that anxiety does not uniformly impede learning under differentiated support conditions. Cross-lagged analysis revealed exploratory temporal associations between self-efficacy and CT performance (SE→CT: β = .40–.43; CT→SE: β = .37–.40). Strategic navigation was the strongest CT success correlate (OR = 3.87). Forum network density tripled (0.23 to 0.68), consistent with the emergence of reciprocal peer communities. Findings are hypothesis-generating and should not be generalized beyond this single-site, small-sample context.