Effect of a Reflection-Guided Visualized Mindtool Strategy for Improving Students' Learning Performance and Behaviors in Computational Thinking Development

Computational thinking (CT) is an imperative competency in the 21st century. Mindtools can assist students in understanding concepts and decomposing tasks during CT development through programming. However, students may encounter challenges in complex CT problem-solving tasks due to being confused when using mindtools without proper guidance. Research evidence shows the potential of reflection in complex CT problem-solving by regulating cognitive activities. Accordingly, this study designed a reflection-guided visualized mindtool strategy to address CT development challenges. A quasi-experiment and lag sequential analysis were conducted by recruiting 97 junior high school students to examine the effects of the proposed strategy on CT development and to explore students' behavior patterns. Results revealed that the proposed approach improved students' CT performance, CT disposition, meta-cognitive awareness, and learning motivation. Students learning with the proposed strategy exhibited more key behaviors of facilitating CT problem-solving (e.g., generalizing the knowledge, re-designing the algorithm scheme, and evaluating the feasibility of their proposed schemes) than students in the control group, revealing the essential process of CT development and enlightening teachers on guiding students to produce such learning processes when cultivating CT.