Abstract: Problem-Based Learning (PBL) is a popular approach to instruction that
supports students to get hands-on training by solving problems. Question Pool
websites (QPs) such as LeetCode, Code Chef, and Math Playground help PBL by
supplying authentic, diverse, and contextualized questions to students.
Nonetheless, empirical findings suggest that 40% to 80% of students registered
in QPs drop out in less than two months. This research is the first attempt to
understand and predict student dropouts from QPs via exploiting students'
engagement moods. Adopting a data-driven approach, we identify five different
engagement moods for QP students, which are namely challenge-seeker,
subject-seeker, interest-seeker, joy-seeker, and non-seeker. We find that
students have collective preferences for answering questions in each engagement
mood, and deviation from those preferences increases their probability of
dropping out significantly. Last but not least, this paper contributes by
introducing a new hybrid machine learning model (we call Dropout-Plus) for
predicting student dropouts in QPs. The test results on a popular QP in China,
with nearly 10K students, show that Dropout-Plus can exceed the rival
algorithms' dropout prediction performance in terms of accuracy, F1-measure,
and AUC. We wrap up our work by giving some design suggestions to QP managers
and online learning professionals to reduce their student dropouts.