Analysis of Repeat Learners in Computer Science MOOCs

Computation is becoming an integral part of education, especially at college-level. Two MOOC courses provide an introduction to computation and are offered by MIT on EdX: 6.00.1x Introduction to Computer Science Programming in Python has run 16 times and 6.00.2x Introduction to Computational Thinking and Data Science has run 11 times. These courses frame the world around computation, and show learners that they can think about problems they see in everyday life in the context of computation. Our paper shows an analysis of repeat learners (learners who enroll in the same course multiple times) and their behaviors over the many runs of these courses. Around 20% of learners in any given run of a course are repeat learners. Of these, the majority are two-time repeat learners (learners who took the same course exactly twice). We show that learners tend to perform better when they retake a course, and especially when they retake the course sooner rather than later. We also look at a subgroup of learners we call repeat cross-referencers (learner who accessed at least two previous runs of the course during the run time of a later run of the course). We found that repeat cross-referencers complete the course at a suspiciously high rate, and we speculate it is because they are looking back at answers from a previous course run. Lastly, we look at how learners perform in an introductory class and an advanced class. We found that many learners who take both courses are more likely to complete both courses and have more active days in the courses than those learners who only do one of the two classes.