Adaptive spaced retrieval practice produced higher end-of-semester posttest performance than learner-directed AI study, but fixed spaced retrieval did not significantly outperform learner-directed study¶

Claim · All claims
Evidence · 1 study · causal r2 · q3 · i2 medium
Currency · newest study 2026 · checked against its sources 2026-10-20 (automated, 1 of 1 entry) · not reviewed by a person

Subclaims¶

q3 i2 ANCOVA revealed a significant effect of condition on posttest performance; G1 significantly outperformed G3, while G1–G2 and G2–G3 differences did not reach the adjusted significance threshold. → Mahir Akgun and Sacip Toker 2026

Evidence¶

Mahir Akgun and Sacip Toker 2026¶

Mahir Akgun and Sacip Toker. (2026). Do Gains from Generative AI–Enabled Adaptive Pretesting Persist? Evidence from a Retention Study. https://arxiv.org/abs/2606.22328 doi:10.48550/arxiv.2606.22328

q3 · i2 · causal · r2

Univariate ANCOVA on the seven-week posttest (parallel 14-item multiple-choice forms, 0–100 scale) in the same experiment. The article reports "partialη 2 =.128" and adjusted means of 78.19 (G1), 74.55 (G2), 67.28 (G3).

"ANCOVA also revealed a significant effect of condition on posttest performance,F(2,85) = 6.24,p=.003, partialη 2 =.128. Adjusted means indicated G1 achieved the highest posttest scores (M = 78.19, SE = 2.37), followed by G2 (M = 74.55, SE = 2.33) and G3 (M = 67.28, SE = 2.12)."

Discussion¶

Related Claims¶

• Spaced Retrieval Improves Retention — the general claim this finding bears on, with the studies behind it
• Post-baseline practice condition has a significant multivariate effect on combined posttest performance and observed practice effort after adaptive AI-assisted pretesting — related
• Structured CAI with spaced practice and spaced review produced better recall and retention than unstructured CAI — related
• Spaced retrieval practice outperformed both massed retrieval practice and no-practice control on delayed financial knowledge, while massed and control did not differ — related
• Generative AI-assisted feedback is associated with a statistically significant pretest–posttest improvement in violin students' intonation performance over a four-week period — related
• How genAI is used, not who uses it, predicts whether independent performance declines or improves — related
• Open AI collaboration yielded the highest supported writing quality, while bounded AI support yielded the highest independent no-AI Week 8 performance — related
• Non-layered AI feedback produced significantly higher post-test performance than layered AI feedback after controlling for pre-test — related
• Adaptive fading with periodic reinforcement yields higher independence than fixed-high or linear-decay scaffolding in simulation — related
• Educational YouTube videos about ChatGPT fall into three structurally distinct discourse groups: learning-oriented scaffolding (G1), skill-building retrieval practice (G2), and output/productivity orientation (G3) — related
• Exploratory ANCOVA controlling for mastery probability favored the adaptive condition, but is not interpretable as a causal treatment effect — related