Abstract
Improving graduation rates is a crucial goal in higher education, but understanding who is at risk of dropping out is resource intensive, creating barriers to effective retention interventions. Thus, identifying factors that predict drop out risk is increasingly of interest, yet current approaches often lack transparency and accessibility. To aid in this effort, DataKind collaborated with John Jay College of Criminal Justice (JJC) to co-create a data-assisted student success tool, utilizing machine learning to identify students at risk of drop out and centering the role of advisors in supporting students. The insights from this tool are used by JJC to provide proactive interventions (i.e., increased academic counseling) for the students identified as most in need of support. After implementing the data-assisted student success tool and associated inventions, JJC saw a rise in senior graduation rates, suggesting the value collaborative and human-centered data science tools have in fostering student success.
Recommended Citation
Harnisher, J., Villanueva, S., Prieto, D., Anzaldua, R., & Beem, K. (2024). Collaborative Development of Machine Learning Algorithms for Student Success at John Jay College. Journal of Access, Retention, and Inclusion in Higher Education, 7(1). Retrieved from https://digitalcommons.wcupa.edu/jarihe/vol7/iss1/2