Using Unbiased Recursive Partitioning to Plan Curricula, Predict Student Outcomes and Repair Equity Gaps (66301)

Session Information: Assessment Theories & Methodologies
Session Chair: Samantha Thompson

Friday, January 6, 2023 (15:50)
Session: Session 5
Room: 323A
Presentation Type:Oral Presentation

All presentation times are UTC-10 (Pacific/Honolulu)

This study focuses on the relationship between end-of-term cumulative grade-point average (GPA) and intrapersonal competencies such as self-compassion, sense of belonging and psychological well-being, for diverse groups of first-time university students. Intrapersonal competencies have been shown to correlate with institutional performance indicators of student success such as GPA, persistence and academic probation. An interdisciplinary team of researchers at a large US public university assembled measures of intrapersonal competency into an online survey, and collected responses from 3,725 students prior to and at the end of the first semester. Student demographic variables were merged with survey results, and unbiased conditional inference trees, a non-parametric machine learning method, were employed for partitioning and predictive purposes. Results of the post-survey revealed significant differences in grade point averages, when students were partitioned into homogenous groups. Perceived stress was a significant factor predicting GPA for all students, and especially for students of color and for those who qualified for federal aid. Sense of belonging measures such as classroom comfort and perceived peer support varied in their prediction of GPA, by sex and socio-economic background. Facets of psychological well-being such as having a sense of autonomy and a sense of purpose in life were also predictive of GPA, depending on level of perceived stress, and whether a student was living on campus or commuting. Overall, this analysis provided decision trees that could be interpreted relatively easily for intersecting student populations, informing the need for population-specific seminar sections with tailored curricula in support of closing equity gaps.

Authors:
Sandra Kahn, San Diego State University, United States
Marilee Bresciani Ludvik, University of Texas at Arlington, United States
Stephen Schellenberg, San Diego State University, United States


About the Presenter(s)
Ms Sandra Kahn is a University Doctoral Student at San Diego State University, United States

See this presentation on the full schedule Schedule



Conference Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Powered by WP LinkPress

Share this Presentation

Posted by Clive Staples Lewis

Last updated: 2023-02-23 23:45:00