Post-study attitudes and self-reports align with annotated behavior patterns: limited-engagement users show rising overreliance concern and less selective AI use¶

Claim · All claims
Evidence · 1 study (2 entries) · causal r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Concerns about overreliance increased more among limited engagement participants than among proactive and critical participants, and inaccurate-response concerns declined among experimental participants. → Jingwei Yi 2026 q2 i? Proactive and critical participants more often reported using AI when stuck; limited engagement participants more often reported using AI throughout learning or only at early stages. → Jingwei Yi 2026 (2)

Evidence¶

Jingwei Yi 2026¶

Jingwei Yi, Yueqi Xie, Jiyan He, Rui Ye, Junming Huang, Bin Zhu, Sean Rintel, Yu Xie, Xing Xie, and Fangzhao Wu. (2026). Learning behavior accounts for background-related advantage in AI-assisted education. https://arxiv.org/abs/2607.10101 doi:10.48550/arxiv.2607.10101

q2 · i? · causal · r2

Pre-to-post survey analysis of perceived risks by course, group, and risk dimension in the 318-participant trial; the article reports overreliance concerns "increase more among limited engagement participants than among proactive and critical participants."

"Concerns about inaccurate AI responses decline more clearly among both limited engagement and proactive and critical experimental participants relative to the control group. This decline suggests that direct experience with AI assistance may increase confidence in response accuracy or usefulness. Concerns about overreliance increase more among limited engagement participants than among proactive and critical participants."

Jingwei Yi 2026 (2)¶

Jingwei Yi, Yueqi Xie, Jiyan He, Rui Ye, Junming Huang, Bin Zhu, Sean Rintel, Yu Xie, Xing Xie, and Fangzhao Wu. (2026). Learning behavior accounts for background-related advantage in AI-assisted education. https://arxiv.org/abs/2607.10101

q2 · i? · causal · r2

Self-reported AI integration methods among experimental participants (Figure 4c, Wilson 95% confidence intervals); the article reports proactive-critical participants "more often report using AI when stuck."

"Across both courses, proactive and critical participants more often report using AI when stuck, suggesting that AI assistance is used as targeted support after independent effort. Limited engagement participants more often report using AI throughout learning or only at early stages, suggesting less selective use of AI assistance during assignment work."

Discussion¶

Related Claims¶

• Among GPT-assisted students, proactive and critical engagement patterns are associated with higher exam scores while limited engagement shows little benefit relative to control — related
• Reviewed studies consistently raise concerns about AI accuracy, overreliance, academic integrity, and reduced critical engagement in Arab EL2 contexts — related
• Student-facing AI raises serious wellbeing, equity, and safety concerns, and some educators express more confidence in teacher-facing tools — related
• University students held positive attitudes toward GAI in DST creation but expressed moderate concern about over-reliance and low concern about reduced social interaction — related
• Tutors reason differently about effort versus progress goals in ways that mirror the target-specific feature patterns, and trust clearly defined feature-backed explanations — related