Teaching Data Reproducibility through Service Learning

Abstract

Data reproducibility is becoming increasingly important in the social sciences, but it has yet to be incorporated into many undergraduate sociology programs. This note describes a service-learning activity that can be added to an introductory statistics course. Students partner with a nonprofit and analyze quantitative data to answer questions selected by the agency. Reproducibility is the central mechanism of communication between the nonprofit, the students, and the course instructor. An assessment of the project suggests that students achieve an understanding of how to create reproducible data. They also come to see its value as a method of communication about data decisions.