Eliminating self-selection: Using data science for authentic undergraduate research in a first-year introductory course

TL;DR

This paper presents an inclusive data science research experience integrated into a first-year course, enabling all students to participate in authentic research without selection barriers, thereby enhancing engagement and learning in STEM.

Contribution

The study introduces a scalable, inclusive research model that eliminates self-selection bias, allowing all students to engage in authentic research early in their academic journey.

Findings

Students analyze large text datasets using discovery tools.

Students identify patterns in congressional speeches.

All students participate without selection barriers.

Abstract

Research experience and mentoring has been identified as an effective intervention for increasing student engagement and retention in the STEM fields, with high impact on students from undeserved populations. However, one-on-one mentoring is limited by the number of available faculty, and in certain cases also by the availability of funding for stipend. One-on-one mentoring is further limited by the selection and self-selection of students. Since research positions are often competitive, they are often taken by the best-performing students. More importantly, many students who do not see themselves as the top students of their class, or do not identify themselves as researchers might not apply, and that self selection can have the highest impact on non-traditional students. To address the obstacles of scalability, selection, and self-selection, we designed a data science research experience for undergraduates as part of an introductory computer science course. Through the intervention, the students are exposed to authentic research as early as their first semester. The intervention is inclusive in the sense that all students registered to the course participate in the research, with no process of selection or self-selection. The research is focused on analytics of large text databases. Using discovery-enabling software tools, the students analyze a corpus of congressional speeches, and identify patterns of differences between democratic speeches and republican speeches, differences between speeches for and against certain bills, and differences between speeches about bills that passed and bills that did not pass. In the beginning of the research experience all student follow the same protocol and use the same data, and then each group of students work on their own research project as part of their final project of the course.