Finding Significant p in Coffee or Tea: Mildly Distasteful

Abstract

Students’ preferences have an impact on their behavior, and behaviors can in turn affect student performance. Earlier work has found that students who tend to work earlier in the course or curse more in their source code tend to perform better. But could other types of preferences also affect student performance? In this work, we examine the relationship between student preferences such as preferring coffee over tea, and students’ performance in the course. Our results suggest that certain preferences are related to better overall performance in the course, but only for certain cohorts of students. Indeed, this work provides an example of how easy it is to find statistically significant correlations in educational settings.