Back to search

University of Alabama Birmingham

Background or Experience? Using Logistic Regression to Predict College Retention

Abstract

dc:description.abstract

Tinto, Astin and countless others have researched the retention and attrition of students from college for more than thirty years. However, the six year graduation rate for all first-time full-time freshmen for the 2002 cohort was 57%. This study sought to determine the retention variables that predicted continued enrollment of entering freshmen at a large urban, four-year, public institution. Logistic regression was utilized to analyze the data collected over a four-year period. The population studied was 1,346 first-time full-time freshmen entering fall 2007. The variables chosen for analysis were ACT composite, cumulative GPA and high school GPA, ethnicity, gender, Pell eligibility, unmet financial need, advising, early alert notices, engagement and freshman year experience courses, honors participation, change of major, campus housing, and supplemental instruction. Data were analyzed by year of enrollment through spring 2011. Correlation studies eliminated the threat of multicollinearity. The logistic regression models passed goodness-of-fit tests for Hosmer Lemeshow, Omnibus Test of Coefficients, and Cox and Snell and Nagelkerke. The analyses found that ACT Composite, cumulative GPA, advising, ethnicity, engagement courses, change of major, and supplemental instruction were predictors for retention. In year one, two, three and four each one point raise in GPA increased the likelihood of persistence by 3.99, 3.31, 3.52, and 11.60 times, respectively. In year one and two students who were White were 2.29 times and 1.74 times more likely to persist, respectively. Living on campus and having advising appointments in the first year increased the likelihood of persisting by a factor of 1.46 and 1.21, respectively. Changing major in the first year increased the likelihood of returning by a factor of 4. In the fourth year, each change of major decreased the likelihood of persisting by a factor of .62; having a higher ACT composite score decreased the likelihood of persisting while supplemental instruction sessions increased the likelihood of persisting. Investigative efforts to validate the coding of participation in freshmen year experience courses found large discrepancies between the reported and actual frequency counts reported by the system. A need to audit and correct student information system data related to retention variables was noted.

Degree

thesis:*
Level thesis:degree_level
Dissertation

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Synco, Tracee M
Contributors dc:contributor
  • Loucrecia Collins
  • Marilyn Kurata
  • Betty Nelson
  • Marcia O'Neal
  • Margaret Rice

Subjects

dc:subject × 6

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
University of Alabama Birmingham
Base URL
digitalcommons.library.uab.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Synco, Tracee M. Background or Experience? Using Logistic Regression to Predict College Retention. Dissertation thesis, https://digitalcommons.library.uab.edu/etd-collection/3090