instructional scaffolding

Progressive mind maps strengthened programming performance beyond chatbot assistance alone

Xindong Ye, Wenyu Zhang, Yuxin Zhou, Xiaozhi Li, Qiang Zhou

Humanities and Social Sciences Communications

A study of 111 seventh-grade students compared chatbot-supported Python learning with two forms of mind-map scaffolding. The group progressing from partially completed maps to independent construction achieved the strongest programming-test results. Questionnaire-based computational-thinking outcomes also favored the scaffolded conditions, but post-test programming self-efficacy did not differ significantly between groups. The study supports examining the structure around AI help, while intact-class allocation and limited process data constrain causal and transfer claims.