HelpCoach led to greater knowledge retention one week later than pre-task help-seeking training alone¶

Claim · All claims
Evidence · 1 study · causal r2 · q3 · i3 large
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i3 HelpCoach participants showed greater knowledge retention one week after the main session than Baseline participants, with Cohen's d = 1.100. → Hyoungwook Jin 2026

Evidence¶

Hyoungwook Jin 2026¶

Hyoungwook Jin, Weirui Peng, Jieun Han, Q. Vera Liao, and Xu Wang. (2026). HelpCoach: Scaffolding Targeted AI Help-Seeking During Problem-Solving. arXiv. https://arxiv.org/abs/2609.28918 doi:10.48550/arxiv.2609.28918

q3 · i3 · causal · r2

In the between-subjects lab study (n=40), participants repeated the same six-item knowledge test one week later (Stage 9). The article reports greater knowledge retention one week later with a printed effect size of Cohen's d = 1.100 (p = .005).

"HelpCoachalso led to greater knowledge retention one week later ( 𝑝=. 005, Cohen's 𝑑= 1.100), and participants' comments suggested broader benefits for self-regulation and metacognition."

Discussion¶

Related Claims¶

• Effective AI help-seeking is itself a skill unevenly distributed among students, and access to an assistant is not equivalent to effective help-seeking — the general claim this finding bears on, with the studies behind it
• HelpCoach increased spontaneous question specificity during chatbot problem-solving compared with pre-task training alone — related
• HelpCoach did not significantly outperform Baseline in improving or retaining self-reported self-regulation — related
• Evidence suggests over-reliance on generative AI may negatively affect students' self-regulated learning behaviors, potentially harming retention and transfer to novel contexts — reports the opposite