Harmful Study Habits in Online Learning Environments with Automatic Assessment

Learning activities are increasingly performed in online learning environments. Automatic assessment is used in many systems to give students immediate and personalized feedback allowing them to solve exercises regardless of time and place. However, this creates new challenges for students because they are required to independently regulate their own learning. Because of the lack of direct guidance, it is easy for students to resort to bad study habits. In this paper, we investigate what kinds of harmful study habits can be observed in online learning environments that utilize automatic assessment. We study data from two computer science courses with programming exercises and algorithm simulation exercises. The results are in line with earlier findings that starting to study near the deadline is linked to inferior performance. This is also observed in a within-subject comparison which suggests that the relationship is causal. Furthermore, signs of trial-and-error problem solving are observed in some students and they are also correlated with inferior performance in the exercises and the exam.