Equity, Inclusion, and Belonging in Undergraduate Data Science

This symposium features four ongoing research papers as part of "Improving Undergraduate Data Science Education at Scale" (NSF grant #1915714). After presenting a brief overview of the project, each author will spend 7-10 minutes presenting their paper and preliminary findings. We will spend the final 20-30 minutes of the symposium facilitating an audience discussion of implications for the field of STEM education and soliciting feedback for the development of our manuscripts. See abstracts of each project below, by author.

Daniel Lobo, "Pathways to Belonging: Peer-to-Peer Effects of Identity-based Interventions in Data Science."

Student-led collaboration, rather than instructor-led instruction, has become a central approach for promoting STEM persistence and achievement, especially for underrepresented students. But how does inequality in student backgrounds impact the sense of belonging of minority students in the highly collaborative and interdisciplinary STEM field of data science? Drawing on survey data and qualitative data from focus groups with undergraduates, we outline how peer-to-peer interactions in a large introductory data science course may reduce minoritized students' perceived sense of belonging, an important individual-level predictor of student achievement. Conversely, we trace the mechanisms through which identity-based affinity group settings improve the sense of belonging of minoritized students.

Michael Ruiz, "Toward a conceptual framework for developing and analyzing inclusive STEM programs in higher education."

The culture of higher education has contributed to gender, racial, and social class achievement gaps in STEM. Findings from social psychology highlight how individuals with such identities are likely to contend with psychological threats pertaining to these identities which then lead these folks to feel less connected to higher education, negatively impact their learning experience, and contribute to leaving STEM related fields. Moreover, recent research has shown how faculty, student instructors, and peer groups can mitigate these threats, resulting in more positive outcomes for these students.

Erin Manalo-Pedro, "Microaggressions and microaffirmations among women of color in undergraduate data science."

To counter the invisibility of undergraduate women of color in data science, we focus on their everyday experiences of struggle and resistance. We analyze qualitative data from focus groups with women of color undergraduates in data science. Through the critical race lenses of microaggressions and microaffirmations, we demonstrate the mundane yet pervasive subtle cues that shape their trajectories. Our preliminary findings contrast moments of exclusion and validation in the overall campus climate, upon entry into data science, through utilization of academic services, and development of career aspirations. We encourage changemakers to acknowledge how daily interactions perpetuate exclusionary norms.

Renee Starowicz, "A theory of near-peer instruction in undergraduate data science."

The opportunity for undergraduate students to serve as instructional staff within a hierarchical training model offers new perspectives and critical reflections on both access and training within higher education. This paper theorizes the ways in which course students and undergraduate instructors understand their role and responsibilities in supporting underrepresented students in a large introductory data science course through a Critical Disability Studies lens. We offer a thematic analysis elucidated via vignettes to inform higher education instructional models, uplift concrete practices that support interdependent sense making and provide for the development of improving peer instruction system-level supports.