Abstract:

Good study and time management skills are an integral part of successful learning. Taking pauses and spacing out learned content over time helps assimilate knowledge and improves long term retention of information. At the same time, spacing out learned content over time does not work well if the pauses are only short intervals. In this study, we explore short pauses and micropauses in learning and their effect on course outcomes. We quantify pause lengths that are considered harmful, whilst also identifying potentially beneficial pause durations.

The data for the experiment comes from an introductory programming course organized during fall 2014. The participants (n=156) in the study used an instrumented programming environment that stored fine-grained data from the learning process, including timestamps of each key press as well as details on the assignments that the participants were working on. In addition, background information on the participants, as well as how they fared in a pen-and-paper course exam was acquired. In the analysis, we quantified the participants working behavior based on the lengths of pauses that they take during working, focusing on the proportionate amounts of pauses of certain lengths, and contrast those to the course outcomes. The results were then validated using data from a separate course that was organized during spring 2015 with (n=47) participants.

Our results indicate that those who continuously work on the task at hand without significant pauses perform well, even when the work may not immediately lead to expected outcomes. Participants who take either a relatively large amount of micropauses (that is, pauses between 5 seconds and 1 minute) or short pauses (between 1 and 5 minutes long) perform considerably poorer than others. It seems that what is done during the pause does not matter -- participants may e.g. be looking at course materials or non-relevant content on the internet.

The results provide evidence that the worst approach to studying is to continuously switch tasks, and supports the wider literature on cognitive load theory and task switching as well as spacing effect. Based on our observations, the most beneficial strategy regarding pauses is to focus on one single thing at a time, and if facing a challenge, fighting it out. It would seem that switching between the primary task at hand (such as working on a programming assignment) and a supporting task (such as reading course material) can be harmful if the participant constantly changes between these tasks, spending only very short amounts of time on one before switching again.

More broadly, this work suggests that the design of learning activities and environments should attempt to minimize short distractions. In an electronic environment, this would likely entail automatically detecting and correcting this kind of harmful behaviour detailed above and structuring the learning activity so that the participants focus on only one cognitively demanding activity at a time. Activities should be designed to avoid situations where a participant is (near)simultaneously completing assignments and browsing course material. Instead of just quickly checking for a fact from the material, it is more beneficial to refresh a longer section of the material.