Truace
Subscribe
70/100

Moderate evidence

GainPublished recordEducationGood Space

Full story
More
Export record JSONEmailLinkedInX

Do AI Writing Tools Improve L2 Writing Performance? A Meta-Analysis of Experimental Evidence, Moderators, and Methodological Quality

Truvace record
TRV-2026-2749raw data (JSON)
Version
v1Published
First published
Evidence type
Systematic review or meta-analysis
Sources
1 sourcePeer-reviewed
sha256
00fb0a4e39db…check it

Artificial intelligence (AI) writing tools—including large language model (LLM)-based systems, automated writing evaluation (AWE) platforms, and AI-powered writing assistants—have become an increasingly prominent feature of second language (L2) writing pedagogy. Yet whether these tools reliably improve L2 writing performance, and under what instructional conditions, remains an open question. This study reports a systematic meta-analysis of experimental and quasi-experimental evidence on the effects of AI writing…

The gain

A meta-analysis of experimental studies published 2021 to 2026 found AI writing tools improved second-language writing performance, with a pooled effect of g = 0.714 (95% CI 0.592 to 0.836) across 22 verified effect sizes.

What this doesn’t fix

Moderator analyses were exploratory, and many included studies lacked delayed transfer measures and transparent effect-size reporting, so lasting effects are not established.

Sources

1 on file
  1. Journal of Educational Technology and Innovationdoi:10.61414/awr9xk46
    Peer-reviewed2026-09-29

What the review found when it combined the studies

These are the review’s own pooled results, copied from its abstract. Truvace did not recalculate them.

  • Second language writing performance with AI writing tools versus comparison conditions (verified subset)

    Hedges' g 0.714 (95% CI 0.592 to 0.836)

    22 studies

    The review’s own words
    The verified model produced a pooled effect of g = 0.714, 95% CI [0.592, 0.836], with low observed heterogeneity (I² = 24.3%).

Code checks every number above against the review’s text before it is shown. The short labels are an AI model’s reading of that text, so the review’s own sentence sits under each one. A pooled figure is only as strong as the studies behind it. How Truvace uses pooled results

Cite this record

v1

Truvace Impact Record TRV-2026-2749, v1: “Do AI Writing Tools Improve L2 Writing Performance? A Meta-Analysis of Experimental Evidence, Moderators, and Methodological Quality.” Truvace, 2026-11-12. /record/TRV-2026-2749 (accessed on the day you copy it). sha256 00fb0a4e39db432f…

Change history

1 version

Every change made to this record since it was first published, and why. Nothing has changed since it went up.

  1. Publishedv100fb0a4e39db…

    Certified into the record

Check this record hasn’t been altered

Check this record hasn’t been altered
How to check it yourself, without trusting this page

Download any version from /api/record/TRV-2026-2749 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.