The effect of feedback on metacognition - A randomized experiment using polling technology

This study explores the effects of formative feedback on students' metacognitive skills when using feedback strategies with polling technology. Using a randomized field experiment among 633 physics students in six schools in Dutch secondary education, we study assessments with the polling technology Socrative, by dividing students into three groups. Students in the cooperative group use a combination of peer discussions and teacher feedback, while students in the individual group use teacher feedback. To compare differences in metacognitive skills, students in the control group only use Socrative, but do not receive formative feedback from either teacher or peers. The results show that there is a significant positive effect of the cooperative treatment on both metacognitive skills and motivation in comparison with the control group. We find that students with low metacognitive skills benefit significantly more from the cooperative treatment than students with high metacognitive skills. No effects are found for the individual treatment. However, girls significantly increase their metacognitive skills and are more motivated than boys, when using an individual treatment. Additionally, a mediation analysis shows that motivation partially mediates the cooperative treatment and metacognitive skills. Based on these results, we recommend a combination of peer discussions and teacher feedback in physics courses.