USING INSTITUTIONAL ANALYSIS OF GRADES TO ASSESS EQUITY OF OUTCOMES IN ONLINE AND IN-PERSON SCIENCE COURSES

We use institutional data (course grades) to examine the extent to which grade-based inequities exist in introductory science courses, online and in-person, related to gender, race/ethnicity, college generational status, and Pell grant eligibility. Related to interests in improving diversity in the sciences, the Howard Hughes Medical Institute’s Inclusive Excellence program is supporting efforts to improve equity in science education. At its core, equity in education means that student course performance, on average, should not depend on demographic group. In a multi-institutional study, Matz et al., (2017) examined ~1M student course enrollments in foundational STEM courses and found gendered performance differences where women earned lower grades in lecture courses and earned roughly equal grades to men in lab courses. Our study extends this prior work by including other demographic categories beyond gender and by including fully-online courses.

For the in-person courses, our results largely conform to those of Matz et al. We find lecture courses, which tend to award lower average grades than the labs, to favor men and labs to favor women. We find students from historically underrepresented minorities (URM) perform worse than white students to a larger degree in the lecture courses and the reverse for labs. Similarly first-generation college students and Pell eligible students follow the same pattern as women and URM students. We find the online course data to obey less clear trends than the in-person course data. For example, while some online courses do award higher grades to the historically advantaged groups, the relationships between inequity and course difficulty are not consistent. Moreover, lab and lecture courses are not so cleanly distinct in terms of average grades. Large online introductory geology and astronomy courses behave somewhat differently in this analysis, offering an interesting case study. The online results have a few possible explanations. Online courses often have low structure and online labs may use exams more than in-person labs, both of which have been shown to disadvantage women and URM students in science. Part of our future work will be to collect course syllabus data in order to test these explanations.