STEM Enrollment Decision Trees as Graduation Predictors for Community College Students Enrolled in Remedial Mathematics.

This article investigates how STEM coursetaking patterns predict graduation outcomes for community college students initially enrolled in remedial mathematics, using decision tree analysis guided by Tinto's academic and social integration framework. Analyzing data from 5,065 students at a suburban community college, the study identified nine decision rules indicating that passing College-Level Mathematics courses (including Algebra II, Statistics, and Precalculus) is the strongest predictor of graduation within two years. Science courses such as Astronomy, Environmental Science, Geology, Marine Biology, Biology, and Anatomy and Physiology also contributed to predicting graduation, suggesting that structured STEM course pathways can support academic integration and persistence. The findings imply that community college advisors and administrators might improve graduation rates by using decision trees to guide students through clear, evidence-based STEM course sequences, particularly for those starting in remedial mathematics. Limitations include the single-institution sample, lack of transfer data, and pre-pandemic context, indicating a need for further research on decision tree advisement tools and their impact on student integration and success.