2025 ASEE Annual Conference & Exposition

This NSF/IUSE-funded research project implemented a research-practice partnership to integrate data science across six undergraduate STEM courses at three diverse universities: a public institution, a private university, and a Historically Black College and University. The project, involving an educational evaluation team and 15+ interdisciplinary faculty and graduate students, developed and implemented 12 discipline-specific data science modules multiple times into the courses, reaching over a thousand students. The project aimed to embed data science learning content within existing STEM curricula to enhance student preparedness for data-driven careers. The modules utilized data from various sources including two labs that produce high-frequency, real-world environmental and traffic data nested within two participating universities.

The project team analyzed the perspectives of students and faculty using a mixed methods approach. Data collection encompassed instructor interviews, student self-assessment collected through online surveys, pre and post module implementation data, and student grades on modules-related assessments. Instructors reported that autonomy in data science content development allowed them to tailor modules to student needs and course constraints, focusing on real-world data and discipline-specific case studies with hands-on exercises. They emphasized data visualization and statistical analysis over other data science topics. Challenges instructors’ faced primarily stemmed from varied student backgrounds and COVID-19-related online teaching adaptations. Student survey analysis revealed increases in self-assessed interest, motivation, confidence, and skills related to data science after one or more module completion. There was general alignment between student self-assessments and instructor evaluations at the course level. Students identified challenges such as difficulty with specific data science topics and tools, while appreciating real-world applications of data science topics and hands-on learning. They suggested more scaffolding and additional time for complex topics such as understanding sensors or using different data analysis software. The synthesis of instructor and student perspectives highlights the effectiveness of integrating tailored, discipline-specific data science content into existing STEM courses. This approach addresses both instructor and student needs, offering a sustainable model for data science education beyond the project's funding period, potentially informing broader efforts to enhance data literacy across diverse STEM disciplines. Details of project work are available at this site: ds4STEM.org