Redefining student success: Applying different multinomial regression techniques for the study of student graduation across institutions of higher education

Daniel R. Jones-White, Peter M. Radcliffe, Ronald L. Huesman, John P. Kellogg

Research output: Contribution to journalArticlepeer-review

29 Scopus citations

Abstract

Current definitions of retention and graduation rates distort the picture of student success by limiting it to completion of a degree at the institution of entry. By incorporating data from the National Student Clearinghouse (NSC), a clearer picture emerges. The NSC data captures retention and graduation at both entry and transfer institutions. To accommodate this polychotomous definition of success, more sophisticated methods of modeling limited dependent variables are needed. Though multinomial logit is often considered the most accessible method, the strict assumptions it imposes may be inappropriate. We therefore compare multinomial regression techniques to assess their utility in modeling multi-institutional student success.

Original languageEnglish (US)
Pages (from-to)154-174
Number of pages21
JournalResearch in Higher Education
Volume51
Issue number2
DOIs
StatePublished - Jan 2010

Keywords

  • Categorical dependent variable
  • Multinomial logit
  • Multinomial probit
  • National Student Clearinghouse
  • Student persistence
  • Student success

Fingerprint

Dive into the research topics of 'Redefining student success: Applying different multinomial regression techniques for the study of student graduation across institutions of higher education'. Together they form a unique fingerprint.

Cite this