ABSTRACT
This study investigated the relationship between depth of processing (DoP) of written corrective feedback (WCF) and the accuracy of the texts written by 46 first-year undergraduates. In a pretest-intervention-posttest design with two intervention groups and one control group, the participants completed a three-stage (writing/WCF processing/revision), time-compressed writing task. Data sources included written texts (writing and revision) and languaging data (provided during the revision stage). We found that (a) WCF is not automatically processed and understood even when learners are made to language about their own errors, and (b) the mere provision of WCF does not guarantee accurate correction if learners do not understand both errors and corrections. Especially relevant are both learners’ DoP of the input (WCF) received and their decision making as to what to do with such input (via corrections or via deletion of the original errors and corresponding reformulation of their original intended meaning).
