The retention gap persisted at 48 hours with no differential forgetting between groups (39% vs 52%)¶
Claim · All claims
Evidence · 1 study · causalr2·q3·i2medium
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person
Subclaims¶
Reading the evidence codes
q is the design tier, i the impact, n the sample, and the kind with its r
the rigour judged within that kind: what each code means.
This claim: no standardized effect is recorded in observations/ for it yet, so no benchmark reading is possible.
q3 i2 AI-Assisted students scored lower on the 48-hour delayed retention quiz (M=0.39 vs M=0.52, Hedges' g = 0.77), and a mixed ANOVA found no group×time interaction, indicating no difference in the rate of forgetting. → Bergh 2026
Evidence¶
Bergh 2026¶
Bergh, C., Tag, B., Vassar, A., & Renzella, J. (2026). Your Programming Students'Cognition with ChatGPT: Higher Performance, Lower Retention, and Reduced Ownership. https://arxiv.org/abs/2609.21194 doi:10.48550/arxiv.2609.21194
q3 · i2 · causal · r2
Delayed cued-recall quiz administered 48 hours after the session to all participants, using the same instrument and blind three-marker scoring. The No AI group again recalled more, with a printed medium-to-large effect size Hedges' g = 0.77, surviving Holm correction.
"The same pattern was observed 48 hours later (𝑀= 0.52,𝑆𝐷= 0.16vs. 𝑀= 0.39,𝑆𝐷= 0.19for No AI and AI-Assisted respectively), also significant (𝑡(52.9)=2.91,𝑝=.005, 𝑔=0.77,𝑝 Holm =.011)"
Discussion¶
Related Claims¶
- Immediate task success does not establish learning that persists without AI support — the general claim this finding bears on, with the studies behind it
- ChatGPT-assisted students scored substantially higher on the coding assessment than No AI students (89% vs 69%) — related
- Review reports a randomized trial in which unrestricted ChatGPT study produced lower delayed retention than traditional methods — related
- In Experiment 3, individual learners showed greater forgetting over 24 hours than paired learners — related
- The retrieval (TTST) advantage appeared at both short (5 min) and long (24 h) retention delays, with no reliable evidence that it differed between them — related
- Learning should be assessed with retention and transfer tests rather than current practice performance — related
Page metadata
| Field | Value |
|---|---|
type |
claim |
id |
chatgpt-lower-48h-retention-no-differential-forgetting |
title |
The retention gap persisted at 48 hours with no differential forgetting between groups (39% vs 52%) |
description |
The retention gap persisted at 48 hours with no differential forgetting between groups (39% vs 52%) |
status |
review |
evidence_strength |
moderate |
newest_evidence |
2026 |
generated |
by process:wiki-ingest on 2026-10-10 |
trust |
unverified — no one has confirmed this page's content |
sources |
1 — bergh-2026 |
Frontmatter is the page's machine-readable record — see the Schema & Guide.