More than Money Matters: An Integrated Model of Graduate Student Persistence

TR Number
Date
2005-04-21
Journal Title
Journal ISSN
Volume Title
Publisher
Virginia Tech
Abstract

Graduate student persistence rarely has been studied, yet it is a very important issue for many higher education constituents—including senior administrators, academic deans, faculty members, students, and families. In light of an alarming statistic that approximately 50% of all graduate students fail to complete their degree, there have been few studies to examine this phenomenon and no new models to explain the relationship between factors that influence graduate student persistence. This dissertation examined persistence by graduate students to degree using data from the National Center for Education Statistics' Baccalaureate and Beyond (B&B:93/97) Longitudinal Survey. First, I proposed a logical model of graduate student persistence using economic, academic, and non-academic variables. Then, I tested the model's ability to predict graduate student persistence to degree. Analyses were conducted to answer the following questions:

  1. Does more than money matter in predicting graduate student persistence?
  2. To what extent do economic, academic and non-academic factors included in the integrated model predict graduate student persistence? a) Does this differ by graduate student status (e.g., master's and doctoral)? If so, how?
  3. Using the integrated model, does the likelihood of persisting in graduate school vary by race/ethnicity? If so, how? a) How do student aid and undergraduate debt influence the observed racial/ethnic group differences in persistence?

The results of this study suggest that more than money matters when predicting graduate student persistence. The integrate model of graduate student persistence, developed for this study, is statistically significantly better at predicting graduate student persistence than a model that only includes economic factors. The integrated model tested to be most fit for predicting persistence among doctoral students. And, finally, using the model, the likelihood of persisting in graduate school varies by race/ethnicity. Student aid tends to improve one's chances of persisting in graduate school while undergraduate debt is associated with failing to persist in graduate school.

Description
Keywords
graduate students, economics, logistic, persistence
Citation