A FEEDBACK SYSTEM TO SUPPORT SELF-REGULATED LEARNING

Detailed feedback specific to individual work is strongly related to improvement in students’ performance. Feedback is a complex phenomenon where students need to construct an understanding of it before it can be used to regulate performance. To develop university students as self-regulated learners, fast and frequent formative feedback is an essential element, as it gives students the opportunity to reflect on their performance, provides information that enables them to improve their work and develop their learning strategies. However there are many obstacles to overcome in both transmitting feedback to students and ensuring that they react in a positive way to the assessment of the strengths and weaknesses of their academic work. We concentrate on two difficulties inherent in providing fast, frequent, formative feedback; external feedback may be interpreted positively or negatively by students, which can impact on their development as self-regulated learners; and the inclusion of detailed feedback increases instructors’ workloads, which can become unmanageable when class sizes become larger.

To ensure that students have opportunities to use the feedback to improve their learning and develop as learners, we have built a computer-mediated system to structure, format, speed-up and de-personalise feedback. The computer-mediated system ensures that all intentional feedback (regarding the quality, correctness, and appropriateness of their work) is delivered in a consistent manner to each student and also mitigates against any unintentional feedback due to incidental social interactions. We provide the students with two types of feedback. A structured descriptive feedback containing information about how the student performed the task, with suggestions on how improvements could be made, and evaluative feedback consisting of a judgement, in the form of a rubric, on the correctness of their work. The descriptive feedback is a combination of automatically generated content (e.g. spelling, grammar, word count) and instructor generated content, which tailors the feedback to each student. This approach allows the instructor to concentrate on higher-level feedback and, by recording the feedback for each student, identify common issues, which speeds-up the feedback process and frees time to develop individual comments for each student.

To examine the effectiveness of our system, we tested it on final year undergraduate students taking a module in statistical consulting. As part of this module students worked on the authentic learning task of writing a conference-style abstract. Our system was used to mediate the feedback they received during this activity. We found the combination of descriptive and evaluative feedback to be well received by the students. The evaluative feedback allowed the students to recognise how they performed in the task and the individualised descriptive comments enabled them to make targeted improvements.