Pseudo Personalised Individual Feedbacks for Large Classes

Individual feedback is extremely important to improve student learning. However, providing individual feedbacks comes at the cost of huge workload which becomes almost impossible to bear in the context of large classes. While the other extreme of providing generic feedback is less time consuming but at the same time has a very little impact on student learning. Therefore, we should look for compromised feedback solutions which could overcome the impracticalities of individual feedback and less effectiveness of generic feedback. We shall present a pseudo personalised feedback mechanism which works on the basis of feedback clustering. We have implemented this idea in which the student work is clustered in different groups according to their correlations. Thus, we provide clustered feedback instead of individual feedback, however, since the feedback provided to a cluster is very similar to an individual feedback due to the high correlations amongst the feedback belonging to one cluster, we call it pseudo personalised individual feedback. We shall present how clustering the feedback can have a stronger impact as compared to generic feedback and consumes much less time as compared to preparing individual feedback