In Education, Delayed Feedback Is Often More Efficient Than Immediate Feedback: A Geometric Explanation

Feedback is important in education. It is commonly believed that immediate feedback is very important. That is why instructors stay often late at night grading students’ assignments – to make sure that the students get their feedback as early as possible. However, surprisingly, experiments show that in many cases, delayed feedback is more efficient that the immediate one. In this paper, we provide a simple geometric explanation of this seemingly counter-intuitive empirical phenomenon. 1 Formulation of the Problem In education, intermediate feedback is useful. Empirical data shows that intermediate feedback helps in education. Namely, the very existence of an intermediate test significantly improves the learning outcomes in comparison with the situation when students only learn about their level of knowledge from the final exam; see, e.g., [2]. In [4], we describe a simple geometric model of learning that explains this improvement – and even explains the percentage by which the learning outcomes improve. Shall feedback be immediate or somewhat delayed? Since the feedback provided right after the test is better for learning that the feedback provided only at the end of the class, it seems reasonable to conjecture that the smaller