Do student-written responses to reflection questions predict persistence and performance in online courses? A text analysis approach

Online learning has been recognized as a promising approach to improve learning outcomes in developing countries where high-quality learning resources are limited. Concomitant with the boom in online learning, there are escalating concerns about academic accountability, specifically student outcomes as measured by persistence and success. This chapter examines whether evidence of reflection found in student written responses to a series of skill-building videos predicts success in online courses. Using a text analysis approach, this study analyzed 1,871 student responses to four reflection questions at a large online university in Panama. A binary logistic regression analysis was conducted to explore whether student persistence was affected by evidence of words associated with significant learning found in student written responses to a set of reflection questions. The results suggest that evidence of words associated with significant learning found in student written responses to reflection questions significantly predicts student persistence in online courses. A Kruskal-Wallis test found median final course grade differences between students who showed no evidence of significant learning in their written responses, and those using 1-13 words associated with significant learning. These results strongly suggest that persistence and performance in online courses are affected by evidence of reflection found in student written responses to reflection questions. These results suggest that a set of reflection tasks assigned early in the course may prove effective in identifying at-risk students.