The Effects of Feedback as Conflict-resolver and Conflict-driver on Learning Outcomes and Error Correction

Abstract

This study investigated the effect of conflict-resolving and conflict-driving feedback on learning and error correction. It also aimed to explore relationship between the feedback types and error correction depending on the learners’ levels of cognitive conflict. In a quasi-experimental study, 105 students from a high school in Gyeong-gi province in South Korea participated. The students were assigned into one of the four types of feedback groups (elaborate, prompted elaborate, example, and prompted example). Using MANCOVA, the effects of feedback types on performance of the immediate and posttests were compared between groups while controlling for pretest scores as covariate. For observation of trends of change of errors at different levels of cognitive conflict between groups, Chi-square test for independence was used. Results of performance on immediate and delayed posttests showed no significant differences between feedback groups. However, a closer observation of the performance on conflict resolution revealed superior effect of prompted elaborate feedback over elaborate and example feedback, and the weakest learning effect of prompted example feedback for high-conflict items. Items of some-conflict did not show difference between feedback groups on error correction in immediate posttest, but only prompted example group scored lowest in delayed posttest. The result indicated that high-conflict errors were overcome most effectively with feedback that aids quick resolution of conflict but which at the same time ensures learners’ cognitive processing of the feedback information. The absence of the testing effect of example feedback supported for reversed testing effect (Pastoetter & Baeuml, 2017). The reason may be attributed to two possible reasons either that: learners’ expectations failed against the type of feedback that demands more interactive cognitive processing, or the feedback functions most sufficiently by providing explanations but with prompts. The result also pointed to the importance of the way prompts should be designed so that learners may not be cognitively overloaded with too demanding cognitive activity with the feedback intervention.