Self-regulation limits supported a targeted teacher role: many students rarely thought before asking AI or could not tell when an AI explanation was wrong¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Nine of 18 students said they rarely or only sometimes thought first before asking AI, and seven of 18 said they rarely or only sometimes could tell when an AI explanation was wrong. → Jin-Guo Liu 2026

Evidence¶

Jin-Guo Liu 2026¶

Jin-Guo Liu, Shang-Qi Lu, Xin-Ran Shi, Long-Li Zheng, Wei Wang. (2026). Test-Driven, AI-Assisted Learning: Replacing Lectures with Weekly Closed-Book Tests. https://arxiv.org/abs/2606.23315 doi:10.48550/arxiv.2606.23315

q2 · i? · design · r2

Self-regulation items from the end-of-term survey (N=18). The authors read these as reasons the teacher role cannot be handed to AI, supporting "a targeted teacher role: diagnosing confusion, marking crucial ideas, and explaining hard concepts". No effect sizes printed.

"Nine of 18 said they rarely or only sometimes thought first before asking AI. Seven of 18 said they rarely or only sometimes could tell when an AI explanation was wrong."

Discussion¶

Related Claims¶

• Students in a 13-week lecture-free course reported accepting the weekly closed-book gate and rated the format as deeper and faster than lectures — related
• Learners misjudge which learning strategies are effective — a broader claim this one bears on
• Most direct SRL promotion targeted cognitive capabilities and rarely mentioned the benefit of strategy use — related
• Students see obtaining explanations for their doubts as ChatGPT's most useful role in learning computer science — related