Publication Details

Abstract

The purpose of this paper is to explore differences in the reliabilities of cumu­lative college grade point averages (GPAs), estimated for unweighted and weighted, one-semester, 1-year, 2-year, and 4-year GPAs. Using cumulative GPAs for a freshman class at a major university, we estimate internal consis­tency (coefficient alpha) reliabilities for the several GPAs. We compare these re­liabilities to similar reliabilities found in the literature. Principal findings are that different cumulative GPAs have differ­ent degrees of reliability and that GPA reliability increases at a decreasing rate with number of semesters completed. Understanding these differences in reli­ability has implications for how GPAs are used by institutional researchers in practical as well as theoretical stud­ies. The literature review and methods of the study should be useful to the institutional researcher who under­takes an investigation that involves GPA reliability.

Authors Joe L. Saupe, University of Missouri Mardy T. Eimers, University of Missouri

Copyright © Association for Institutional Research 2013