Designing research into the effects of grammar correction in L2 writing: Not so straightforward

In the research conducted by Truscott and Hsu (2008), the authors demonstrate that although rewriting corrected drafts results in lower grammar error rates on the rewritten texts, this effect does not carry over to a subsequent new writing task. The authors conclude that the result indicates that there may have been no language improvement from the correction-rewrite procedure. This commentary, however, attempts to show that the conclusion seems to be true, but arguably for the wrong reasons because the details in the various sets of data suggest a possible alternative explanation. On the basis of the analysis offered, some general suggestions about the design of future research on the effects of correction in L2 writing are made.