Determining Academic, Background, and Financial Predictors of Community College First Year Retention using Data Mining Techniques
Loading...
Authors
Pace, Camille Gasaway
Issue Date
2021-06
Type
Dissertation
Language
en_US
Keywords
Dissertations, Academic--United States , College dropouts--Prevention , Community colleges , Data mining
Alternative Title
Abstract
Even with extensive retention research dating from the 1960s, community colleges still struggle to identify the reasons why students do not return to college. Data mining has allowed these retention models to evolve to identify new patterns among student populations and variables. The purpose of this study was to create a predictive model for student retention using background, academic, and financial factors serving as a guide for other community colleges to use when investigating institutional retention. Four different data mining models (neural networks, random forest trees, support vector machines, and logistic regression) identified significant factors for retention. The models were compared to identify if one outperformed the others on five different evaluation metrics.
The number of credit hours was consistently the most important variable in retention. In addition, the interactions between the number of credit hours, GPA, and financial aid variables were significant in student retention in their first year. The interaction between GPA, financial aid variables, and the number of remedial hours was also crucial for the first-year retention. There were no consistent variables among the retention models that can predict students' nonretention in the first year of their college career. Many background predictors (age, gender, race, or ethnicity) were not significant in predicting retained or nonretained students. The comparison of the retention models found the random forest model had the best performance for accurately classifying the nonretained and retained students overall and the retained students individually.
Keywords: Retention, Community College, Data Mining, Academic Factors, Background Factors, Financial Factors
Description
Citation
Publisher
License
This dissertation is protected by the Copyright Laws of the United States (Public Law 94-553, revised in 1976). Consistent with fair use as defined in the Copyright Laws, brief quotations from this material are allowed with proper acknowledgement. Use of the materials for financial gain with the author's expressed written permissions is not allowed.