Files and links (1)
PDFOpen Access (License Unspecified), Open Access
Dissertation
Using predictive analytics to forecast success for an adult learner student population: a quantitative study
Doctor of Education (Ed.D.), Drexel University
Sep 2025
DOI:
https://doi.org/10.17918/00011220
Abstract
Adult learners enrolled in online undergraduate programs face unique challenges, including balancing academic work with employment and caregiving responsibilities, navigating financial pressures, and engaging in institutional systems designed primarily for traditional-aged students. These factors contribute to elevated attrition risk and highlight the need for predictive models that more accurately identify which adult learners are at risk of leaving higher education. The purpose of this study was to examine which variables most strongly predict first-to-second-year retention among adult learners at a large, private institution that primarily serves traditional, on-campus undergraduates. Using a quantitative design grounded in a post-positivist paradigm, the study analyzes the demographic, academic, and learning engagement data to build and test predictive models. Logistic regression and survival analysis techniques were applied to pooled, admission year, and admission term cohorts to assess the stability and strength of predictors across different grouping structures and over the four-year degree span. Findings revealed that early academic momentum (credits attempted and earned), grade performance, and stop-out behavior were consistently strong predictors of second-year retention. However, the significance and magnitude of specific predictors varied by modeling cohort structure, demonstrating that pooled, admission year, and admission term models identified different combinations of significant factors. Survival analysis further showed that while early predictors retained influence in later years, their strength diminished over time, suggesting that additional evolving factors contributed to persistence beyond the second year. The results underscore the importance of tailoring predictive systems to adult learners' experience and ensuring that institutional interventions are designed with this population in mind. The findings have practical implications for enrollment management, advising, and student success initiatives in contexts where supporting adult learners is both a moral and strategic imperative.
Metrics
12 File views/ downloads
12 Record Views
Details
- Title
- Using predictive analytics to forecast success for an adult learner student population
- Creators
- Michelle F. Spina
- Contributors
- Harriette Rasmussen (Advisor)
- Awarding Institution
- Drexel University
- Degree Awarded
- Doctor of Education (Ed.D.)
- Publisher
- Drexel University
- Number of pages
- ix, 280 pages
- Resource Type
- Dissertation
- Language
- English
- Academic Unit
- School of Education (1997-2026); Drexel University
- Other Identifier
- 991022138283304721