Background: Instructional transparency makes a course’s learning goals, evaluation criteria, and path to success clear to students, with the goal of improving equity in higher education. Increased transparency may improve equity by bolstering students’ self-efficacy and sense of belonging in computing, both of which are correlated with persistence in the field.

Purpose: We aim to understand whether there are group differences in how students perceive and benefit from instructional transparency. We are additionally interested in understanding whether perceiving instructional transparency can positively influence students’ self-efficacy and sense of belonging and, therefore, contribute to the persistence of students from historically underrepresented groups in computing.

Methods: To investigate these relationships, we used linear regressions to analyze survey responses from 11,046 undergraduate students from 203 institutions.

Findings: We found that there are group differences in students’ perception of transparency in their CS courses: students who identify as women, first-generation college students, and/or disabled reported perceiving less instructional transparency than their peers. We also found that perceiving more transparency has a positive correlation with students’ self-efficacy and sense of belonging in computing while controlling for important confounding variables, such as prior CS experience. We further demonstrated that this relationship is different for certain groups of students: first, for Black students and first-generation college students, perceiving transparency has a larger positive impact on their self-efficacy, and second, for Hispanic students, perceiving transparency has a smaller positive impact on their sense of belonging.

Contributions: Our work constitutes one of the first empirical, multi-institutional investigations of the perceptions and benefits of transparency in CS classrooms that focuses on group differences. Our work also includes a theoretical articulation of the mechanisms through which transparent teaching practices may influence students’ self-efficacy and sense of belonging in computing. Taken together, our empirical findings and theoretical argument provide important evidence for the benefits of instructional transparency in CS courses, particularly as it relates to improving equity in computing.