Students in the AI-integrated database course showed large, statistically significant pre/post gains on seven identical database assessment questions¶

Claim · All claims
Evidence · 1 study (2 entries) · design r2 · q2 · i3 large
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Pre/post assessment scores improved from a mean of 4.25 to 6.83 out of 7, a highly significant increase (t(12)≈5.10, p < .001) with a mean score increase of 2.58 points. → Hosseini 2026 q2 i3 Cohen's d for the learning gains was 1.49, which the authors describe as a large, practically meaningful effect. → Hosseini 2026 (2)

Evidence¶

Hosseini 2026¶

Hosseini, H. (2026). The Pedagogy of AI Mistakes: Fostering Higher-Order Thinking. https://arxiv.org/abs/2605.05472 doi:10.48550/arxiv.2605.05472

q2 · i? · design · r2

Pre/post assessment in an asynchronous undergraduate database design course (n = 13), with seven identical multiple-choice questions administered at course start and embedded in the midterm/final. Mean rose from 4.25 (SD = 1.57) to 6.83 (SD ≈ 0.39); "A paired t-test confirms this improvement is highly significant (t(12)≈5.10, p < .001)". No effect size is printed in this test statistic.

"A paired t-test confirms this improvement is highly significant (t(12)≈5.10, p < .001), with a mean score increase of 2.58 points."

Hosseini 2026 (2)¶

Hosseini, H. (2026). The Pedagogy of AI Mistakes: Fostering Higher-Order Thinking. https://arxiv.org/abs/2605.05472. doi:10.48550/arxiv.2605.05472

q2 · i3 · design · r2

Effect-size analysis of the same pre/post gains in the 13-student course, following a Shapiro–Wilk normality check (W = 0.95, p = 0.64). The authors report "a large effect size (d= 1.49)" and interpret the improvement as practically meaningful.

"We then examined the magnitude of learning gains using Cohen’sd, which yielded a large effect size (d= 1.49)."

Discussion¶

Related Claims¶

• AI-supported instruction has a moderate positive effect on student achievement in STEM education (g = 0.670 under a random effects model) — the general claim this finding bears on, with the studies behind it
• Per-question pre/post analysis showed significant improvement on almost all questions, with Q1 (DBMS use) and Q4 (basic normalization) as exceptions due to high pre-test scores — a narrower finding that bears on this claim
• Paired pre/post AI readiness scores rose descriptively from 82.5% to 89.9% (mean gain 7.4 percentage points) among seven co-creation clients, treated as descriptive context only — related
• College students' scores on an online source credibility assessment improved significantly after four course-embedded modules in an asynchronous nutrition course — related
• Students' overall AI knowledge improved significantly after four integrated AI-biology lessons — related
• The integrated G4L approach combining graph-based modeling, database-driven tracking, and generative content creation successfully supported learner navigation and domain comprehension — related