Exploring potential roles of academic libraries in undergraduate data science education curriculum development

Abstract

Undergraduate data science education is receiving increasing interest in many higher education institutions in the U.S., with the proliferation of data and data related work and research. As an emerging interdisciplinary study field, data science curriculum is typically a collection of individual data science related courses from different schools and departments, most of which are teaching data science in a siloed fashion. Therefore, it is necessary to map the landscape of existing curricula and explore how academic libraries can collaborate and contribute to undergraduate data science education. In this study, we analyzed teaching content and topics of over 100 data science related courses at Purdue University to map the landscape and explore roles of academic libraries to support data science education curriculum. Our results indicate most existing courses focused on ‘hard-core’ scientific analytic principles, such as computer science, statistics, and domain-specific skills. Courses of data-oriented skills, such as data management, data ethics, and data communications were limited across disciplines. In addition, data science courses were more likely targeting STEM students at upper levels (3rd and 4th year students). Academic libraries can enrich data science education efforts, by supporting credit courses, certificate programs, and other co-curricular activities to provide learning opportunities to all students, particularly 1st and 2nd year students and non-STEM majors.