aceModerate evidence
Published recordEducationGood Space
More
Investigating the effect of artificial intelligence in education (AIEd) on learning achievement: A meta-analysis and research synthesis
- Truvace record
- TRV-2026-2842raw data (JSON)
- Version
- v1Published
- First published
- Evidence type
- Systematic review or meta-analysis
- Sources
- 1 sourcePeer-reviewed
- sha256
- 24fd5b9a0bb0…check it
Scant information exists about how AI with its different technologies might affect learning achievement in different educational fields across different educational levels and geographical distributions of students. Closing this gap can therefore help stakeholders understand under which learning conditions artificial intelligence in education (AIEd) might work or not, hence achieving better learning achievement. To address this research gap, this study conducted a meta-analysis and research synthesis of the effe…
The gain
A meta-analysis of 85 quantitative studies with 10,469 students found AI in education had a very large effect on learning achievement (Hedges' g = 1.10), with chatbots showing the largest effects.
What this doesn’t fix
The effect of AI in education was moderated by field of education, level of education, learning mode, intervention duration and geographical distribution, so the pooled result varies across settings.
Sources
1 on file- Information Developmentdoi:10.1177/02666669241304407
Cite this record
v1Truvace Impact Record TRV-2026-2842, v1: “Investigating the effect of artificial intelligence in education (AIEd) on learning achievement: A meta-analysis and research synthesis.” Truvace, 2026-11-12. /record/TRV-2026-2842 (accessed on the day you copy it). sha256 24fd5b9a0bb0cf86…
Change history
1 versionEvery change made to this record since it was first published, and why. Nothing has changed since it went up.
Certified into the record
Check this record hasn’t been altered
How to check it yourself, without trusting this page
Download any version from /api/record/TRV-2026-2842 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.