The Efficacy of Written Corrective Feedback in Improving L2 Written Accuracy: A Meta‐Analysis

21 primary studies published 1980-2013 (Lim & Renandya 2020 cite the same review as k = 24; the discrepancy is unresolved here)

Second-language English learners. The corpus is overwhelmingly post-secondary - a contemporaneous critical review of 42 written-corrective-feedback studies found 93% used adult learners and only about 9.5% ran in elementary, secondary or high schools.

The most-cited quantitative synthesis of written corrective feedback and the standard positive citation in the error-correction debate. Recorded because "does correcting errors improve accuracy" is the closest large literature to the conventions question - and because its population makes it inapplicable to the age band this archive is being built for, which is itself the finding worth storing. Paywalled; the effect size and moderator list here come from the abstract and secondary reports, and the revision-versus-new-writing breakdown could NOT be verified.

Overall g = 0.54 in favour of written corrective feedback on L2 written accuracy. Learner proficiency was the strongest moderator; setting mattered (ESL greater than EFL), as did genre, number of treatment sessions, target language and publication year. Direct versus indirect feedback, and focused versus unfocused feedback, were both NON-SIGNIFICANT - the two distinctions practitioners argue about hardest did not separate. An independent replication of the same literature with different inclusion rules (Lim & Renandya 2020, k = 35) obtained g = 0.59 (95% CI 0.42 to 0.76) and reversed the setting moderator, finding FL 0.70 above SL 0.48, with low-proficiency learners at g = 0.98 against 0.36 for intermediates. Truscott has published critiques arguing the weight in these pooled estimates rests on a few closely related experiments; those critiques are not peer-reviewed and are not recorded as sources.

Low. Pooled experimental comparisons in which learners are assigned to feedback conditions.