aceModerate evidence
More
AI feedback boosts language learners' results, meta-analysis of 55 studies finds
Study title: Exploring the effects of AI-powered feedback in language learning: a meta-analysis
- Evidence type
- Systematic review or meta-analysis
- Sample
- n = 55
- Sources
- 1 sourcePeer-reviewed
- Published
- 1 min read
- Truvace record
- TRV-2026-2751cite or check it
Study title: Exploring the effects of AI-powered feedback in language learning: a meta-analysis
A meta-analysis pooled 55 eligible studies, found through searches of Web of Science and EBSCO covering 2004 to 2024 plus manual journal searches, on AI-generated feedback in language learning. It reported a pooled effect of g = 1.106 on learning performance across 49 studies.
Read the study abstract
Despite growing interest in artificial intelligence in education, the effectiveness of AI-generated feedback in language learning remains unclear. This meta-analysis examined its impact on language learning performance and affective outcomes. A systematic search of the Web of Science and EBSCO databases (2004–2024), supplemented by manual journal searches, identified 55 eligible studies. Results indicated significant positive effects on learning performance (g = 1.106, p < 0.001; k = 49) and affective outcomes (…
The gain
A meta-analysis of 55 studies found AI-generated feedback improved language learning performance (g = 1.106, 49 studies) and affective outcomes such as learner attitudes (g = 0.823, 16 studies).
In brief
For affective outcomes, measured in 16 studies, the pooled effect was g = 0.823. Both effects were statistically significant at p < 0.001.
The size of the effect varied with how the feedback was set up. Intervention duration, language skill, feedback source, delivery mode, target language and learning context all moderated performance results, while generative AI use mattered for affective outcomes.
Main points
- The meta-analysis identified 55 eligible studies from Web of Science and EBSCO, covering 2004 to 2024, plus manual journal searches.
- Learning performance improved with a pooled effect of g = 1.106 (p < 0.001) across 49 studies.
- Affective outcomes improved with a pooled effect of g = 0.823 (p < 0.001) across 16 studies.
- Performance effects differed by intervention duration, language skill, feedback source, delivery mode, target language and language learning context.
- Affective effects differed by use of generative AI, language skill, delivery mode, target language and language learning context.
The full story
The authors set out to clarify whether AI-generated feedback works in language learning, a question they describe as unresolved despite growing interest in AI in education. They searched two databases for studies from 2004 to 2024 and added manual journal searches.
Of the 55 eligible studies, 49 reported learning performance and 16 reported affective outcomes, with significant positive pooled effects for both.
Moderator analyses indicate the benefits are not uniform. Six factors shaped performance outcomes, and a partly different set, including whether generative AI was used, shaped affective outcomes.
Sources
1 on file- Educational Psychologydoi:10.1080/01443410.2026.2735272
Talk it over
Who gets this benefit, and who was left out of the study or story?
AI feedback boosts language learners' results, meta-analysis of 55 studies findshttps://truvace.com/item/exploring-the-effects-of-ai-powered-feedback-in-language-learning-a-meta-analysi
Who gets this benefit, and who was left out of the study or story?
The gain A meta-analysis of 55 studies found AI-generated feedback improved language learning performance (g = 1.106, 49 studies) and affective outcomes such as learner attitudes (g = 0.823, 16 studies).
Answer in the threadHow do we know? What can this kind of evidence show, and what can’t it?
The evidence This rests on a systematic review or meta-analysis with a sample of 55, from Educational Psychology.
Answer in the threadWhat does it leave open? What else would you want to know before acting on it?
Answer in the threadHave you seen AI used this way where you study, work or live? Did it go the same way?
Answer in the thread
The questions are the same on every record. The details under them are taken from this record as written.
What readers say