How am I going? Behavioral engagement mediates the effect of individual feedback on writing performance

Successful feedback should provide learning goals, evaluate current performance and indicate improvement strategies. Furthermore, feedback can only positively affect student performance if students actively engage with it. Thus, it is necessary to consider the feedback reception process in addition to the feedback information itself. This study compares the effects of different types of feedback information on the writing performance of lower secondary students of English as a foreign language (EFL) in a digital learning environment. Behavioral engagement was considered as a mediator of the feedback effect. Participants were N = 338 eighth- and ninth-grade EFL students (54.7% female) enrolled in lower-secondary education within the Swiss school system. We conducted a web-based randomized-controlled experiment, in which students were randomly assigned to four conditions receiving varying amounts of rubric-based feedback information. We used log data (time on feedback page) as a proxy for their behavioral engagement with the feedback. Even though writing performance improved substantially across conditions, there were no differential effects of the type of feedback information on performance. However, EFL learners who received individual performance information spent more time with the feedback, especially those with low prior achievement. Mediation analysis showed that the effectiveness of the feedback was mediated by the time spent on the feedback as an indicator of students' behavioral engagement. Advantages for individual performance feedback over more general information were observed as a function of time spent with the feedback. This finding implies that engagement should be considered in feedback research.