Integrating Data Science Into Undergraduate Science and Engineering Courses: Lessons Learned by Instructors in a Multiuniversity Research-Practice Partnership

Abstract

Contribution: This article discusses a research-practice partnership (RPP) where instructors from six undergraduate courses in three universities developed data science modules tailored to the needs of their respective disciplines, academic levels, and pedagogies. Background: STEM disciplines at universities are incorporating data science topics to meet employer demands for data science-savvy graduates. Integrating these topics into regular course materials can benefit students and instructors. However, instructors encounter challenges in integrating data science instruction into their course schedules. Research Questions: How did instructors from multiple engineering and science disciplines working in an RPP integrate data science into their undergraduate courses? Methodology: A multiple case study approach, with each course as a unit of analysis, was used to identify data science topics and integration approaches. Findings: Instructors designed their modules to meet specific course needs, utilizing them as primary or supplementary learning tools based on their course structure and pedagogy. They selected a subset of discipline-agnostic data science topics, such as generating and interpreting visualizations and conducting basic statistical analyses. Although instructors faced challenges due to varying data science skills of their students, they valued the control they had in integrating data science content into their courses. They were uncertain about whether the modules could be adopted for use by other instructors, specifically by those outside of their discipline, but they all believed the approach for developing and integrating data science could be adapted to student needs in different situations.