Computational thinking as a mediator between modeling self-efficacy and flexible thinking in pre-service science teachers

Introduction: In psychology, constructs such as self-efficacy, cognitive flexibility, and computational thinking (CT) are recognized as mechanisms that explain how individuals adapt to uncertainty and complexity. Self-efficacy regulates both motivation and cognition, while flexible thinking reflects the adaptive capacity to generate alternatives and cope with ambiguity. CT extends beyond technical domains as a cognitive framework that structures processes such as abstraction and reasoning. This study explores the mediating role of CT in the relationship between mathematical modeling self-efficacy and flexible thinking among pre-service science teachers. Methods: Participants were 244 pre-service science teachers from two state universities in Türkiye. Data were gathered using the Mathematical Modeling Self-Efficacy Scale, the Computational Thinking Skills Scale, and the Flexible Thinking in Learning Scale. Confirmatory factor analyses supported validity and reliability. PROCESS Macro (Model 4) tested mediation with controls for gender and grade. Results: = 0.378 when CT was included, indicating partial mediation. Bootstrap analysis confirmed the indirect effect (0.224, 95% CI 0.150, 0.298). Discussion and conclusion: Findings indicate that self-efficacy beliefs influence adaptive cognitive capacities both directly and indirectly through CT. CT functions as a psychological bridge, linking self-efficacy with flexible thinking, and highlights the importance of integrating these constructs in educational psychology to foster resilience, adaptability, and coping with complexity.