The exact text we saved

This is the version exactly as it was saved, character for character. The SHA-256 hash below is a 64-character code worked out from that text. Run the same text through SHA-256 and you get the same code; change a single letter and the code comes out completely different. Each version also stores the previous version’s hash, which chains them together so no version can be swapped out quietly.

TRUVACE RECORD VERSION
record: TRV-2026-2785
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-12T17:23:40.069640Z
status: published
lens: g_space
sector: education
headline: Can Generative AI Feedback Effectively Enhance Learning Outcomes? A Meta-Analysis of 36 Experimental and Quasi-Experimental Studies
dek: Although generative artificial intelligence (GenAI) feedback shows promise for educational applications, its actual impact on learning outcomes and the factors influencing its effectiveness remain unclear. This study conducted a systematic review and meta-analysis to evaluate the effectiveness of GenAI feedback and identify key moderating factors. Following the PRISMA 2020 guidelines, we reviewed 36 experimental and quasi-experimental studies published between 2023 and 2025, yielding 72 effect sizes. The results…
gain_title: A meta-analysis of 36 experimental and quasi-experimental studies (72 effect sizes, 2023 to 2025) found generative AI feedback had a moderate positive effect on academic achievement (g = 0.61).
problem_title: (none)
trace_subject: (none)
gain_reading: A meta-analysis of 36 experimental and quasi-experimental studies (72 effect sizes, 2023 to 2025) found generative AI feedback had a moderate positive effect on academic achievement (g = 0.61).
gain_evidence: GenAI feedback had a moderate positive effect on academic achievement (g = 0.61)
problem_reading: (none)
problem_evidence: (none)
quick_read: A systematic review and meta-analysis following PRISMA 2020 guidelines pooled 36 experimental and quasi-experimental studies published between 2023 and 2025, yielding 72 effect sizes. It found that generative AI feedback had a moderate positive effect on academic achievement (g = 0.61).

The size of the benefit depended on how students were taught. Effects were stronger in learner-centered settings that promote active construction of knowledge than in teacher-centered, receptive instruction. Educational level, discipline, intervention duration and the role the AI played showed no significant moderation.

Benefits were strongest for cognitive outcomes, promising but less established for metacognitive development, and modest for non-cognitive outcomes. The authors say AI feedback should complement constructivist teaching and that teacher emotional support should be preserved.
limitation: Benefits for metacognitive development are described as promising but less established, and effects on non-cognitive outcomes were modest, so the authors call for further research on those outcomes.
tag: Evidence-backed gain
key_points: The review covered 36 experimental and quasi-experimental studies published between 2023 and 2025, producing 72 effect sizes. | The pooled effect on academic achievement was moderate and positive (g = 0.61). | Teaching method significantly moderated results, with stronger effects in learner-centered environments than in teacher-centered instruction. | Educational level, discipline, intervention duration and the AI's role showed no significant moderation. | Cognitive outcomes improved most; metacognitive benefits were promising but less established, and non-cognitive effects were modest. | A three-level random-effects model accounted for dependencies among effect sizes, correcting the underestimated standard errors of conventional two-level models.
rundown: The authors followed PRISMA 2020 guidelines and pooled 36 experimental and quasi-experimental studies published from 2023 to 2025. Because many studies reported several outcomes, the analysis used a three-level random-effects model to handle the 72 dependent effect sizes.

The overall result was a moderate positive effect on academic achievement (g = 0.61). Subgroup analysis showed that learner-centered teaching, where students actively construct knowledge, amplified the effect compared with receptive, teacher-centered instruction.

By outcome type, the largest impact was on cognitive outcomes, with promising but less established gains in metacognitive development and modest effects on non-cognitive outcomes. The authors recommend using AI feedback as complementary scaffolding within constructivist pedagogies and preserving teacher emotional support to foster students' non-cognitive development.
sources:
- peer_reviewed | Education Sciences | https://doi.org/10.3390/educsci16060816 | 2026-05-22
prev: 0000000000000000000000000000000000000000000000000000000000000000

SHA-256 hash

82685272e570eaff49dfe20998d23cc8723876617d59c6409cc62d03a36efb48

Previous version’s hash

0000000000000000000000000000000000000000000000000000000000000000