Need: Becoming a responsible and human-centered data scientist is not only a matter of knowing how to code and reason statistically. Data science for the twenty-first century needs practitioners who understand not only how to analyze data but also how socially constructed categories shape data, how the provenance of data matters, and how data is used to tell stories and change minds. Literature suggests that integrating the humanities into data science education can enhance student learning of essential concepts and foundational reasoning skills, such as those collectively known as “data acumen,” and increase the appeal of STEM courses for a wider range of students, including women and students of color. Our interdisciplinary team from humanities and STEM fields has crafted an introductory course that integrates core humanities skills—close reading, contextual thinking, a critical lens—with traditional statistical and computational approaches to data science education. Guiding Questions: How well does our innovative pedagogical approach cultivate students’ data acumen? In what ways and to what degree does our introductory course change attitudes about STEM, data science, and the humanities? How effectively do students achieve student learning outcomes (SLOs)? Outcomes: Two professors from applied mathematics and English team-taught “Inclusive Interdisciplinary Data Science for All” for the first time in Fall 2021 to 56 students. Using the Team-Based Learning pedagogical strategy, students learned a variety of statistical thinking skills, practiced and applied computing concepts, and implemented data-science-relevant humanities ways of thinking about sources, human motivations, ethics, the power and folly of categorization, and the rhetoric of data visualization. Students collaborated with their teammates to complete team application exercises and assignments, culminating in a final project to find and explore a data set of their own choosing. We found that students’ attitudes remained positive toward STEM, data science, and the humanities. Students self-reported having achieved most of the SLOs, with the outcome of learning to collaborate effectively with teammates being the highest rated. An analysis of the final exam corroborated this finding that the course was effective in helping students learn general data science skills, the epistemology of data science, ethics, communication and rhetoric, statistical thinking, and R programming skills. Based on this first iteration of the course, we have refined the course’s “arc” for providing a humanistic introduction to data science that equips students with fundamental statistics and computing resources they will need for the twenty-first century. Broader Impacts: We are working toward a better understanding of how all undergraduates can be educated in data science to become effective and ethical producers of data-driven inquiry and to act as informed, reflective, and critical consumers of data and research results. We strive to teach 120 undergraduates in Fall 2022 and 600 by 2025. By crafting a more inclusive and human-centered approach to teaching the foundations of data science, and by developing a new collaborative model of data science education that could be adapted nationwide, we hope to positively impact STEM education, leading to a more diverse, creative, and innovative national workforce and a more STEM-literate public.
Nathan Pieplow, University of Colorado Boulder; David Glimp, University of Colorado Boulder; Vilja Hulden, University of Colorado Boulder; Brett Melbourne, University of Colorado Boulder; Jane Garrity, University of Colorado Boulder; Estelle Lindrooth, University of Colorado Boulder; Michael Schneider, University of Colorado Boulder