Student Retention In An Historically Black Institution - Statistical Data Included

College Student Journal, March, 2001 by CLEVE McDANIEL, Steven W. Graham

This study looked at the ability of pre- and early-matriculation variables to predict the retention or attrition of black residential and white commuter students at an historically black, open admissions university. Using a categorical scheme developed by Bean & Metzner (1985) as an initial framework, we examined 25 predictor variables to see if they could predict future retention status. With retention status as the dependent variable, direct and forward stepwise models were developed and comparisons were made between the white commuter students and the black residential students.

The predictor variables that had the highest correlation coefficients with one year retention status were ACT Test Score, ACT math subscore, adequacy of prior education, high school grade point average, high school rank, and student view of actual self. These results corroborate a substantial number of studies that concluded that differences exist on pre- and early-matriculation variables for returning and withdrawing students.

Introduction

It was conservatively estimated that in the academic year 1993 over 800,000 entering freshmen withdrew from post-secondary education institutions in the United States prior to their second year (American College Testing, 1993; Snyder & Hoffman, 1995). Measured in terms of departure, "41 of every 100 entrants will depart the higher educational system without earning a college degree. Most (three-quarters) of them will leave school in the first two years of college, the greatest proportion occurring in the first year of college" (Tinto, 1987, p. 21).

While retention issues associated with a changing student population are relevant to most colleges and universities, they are of particular concern to institutions that have high percentages of nontraditional students. Because nontraditional and minority students tend to have higher withdrawal rates than do traditional-aged students (Astin, 1975), recent shifts in the composition of the higher education student population may cause concern for colleges and universities today. Among all public four-year institutions, the freshman-to-sophomore attrition was highest at those institutions with open admissions policies and at historically black colleges and universities (American College Testing Program, 1993). Consequently, these types of colleges must look for ways to identify students who might be at risk and to assist them, regardless of their level of preparation, in reaching their college goals.

To shed light on this issue, we attempted to develop a statistical model to identify those prone to dropping out based only on pre- and early-matriculation variables. If we could identify factors colleges could evaluate prior to admission, it would help their recruiting and retention strategies. Our goal was to predict the retention status of black residential and white commuter students at an institution with an open admissions policy and to examine the nature of any differences between minority and nonminority students. These findings could assist college officials in their attempts to reduce attrition and identify students most likely to succeed at their schools.

Review of literature

The most widely discussed and researched retention model is one developed by Tinto (1975). In this model, student retention is a function of' a complex series of interactions between a student and the institutional environment. Background characteristics (e.g., family background, individual attributes, and pre-college schooling) interact with each other and, in turn, influence both commitment to the institution and to graduation. Such commitment is increased to the extent that the student/institutional match fosters both academic and social integration. Elements of the Tinto model have been demonstrated to be particularly appropriate for explaining student retention and attrition for the typical student at a residential institution (Munro, 1981; Pascarella & Terenzini, 1979).

However, this model is not as effective in explaining retention for commuter and other non-traditional students (Grosset. 1989). Bean and Metzner (1985) found that "the most important (retention) variables are likely to differ for subgroups such as older students, part-time students, ethnic minorities, women, or academically under-prepared students at different types of institutions ..." (p. 529). In fact, Terenzini and Pascarella (1978) suggested that "the academic and social correlates of attrition may be different for different kinds of students" (p. 364). Even Tinto (1982) indicated concerns about the applicability of his model to non-traditional students and institutions, "it (the model) fails to highlight the important differences in education careers that mark the experiences of students of different gender, race, and social status backgrounds" (p. 689). Such problems with his and other models led Grosset (1989) to conclude "since there is not a generalized, all-purpose attrition model for all institutions, colleges need to develop their own models" (p. 15).