The problems focused on in this study are to determine (1) if racial differences exist when American College Testing Program (ACT) scores are used to predict Grade Point Average (GPA); (2) how placement decisions may be affected if differences do exist; (3) and what guidelines or recommendations can be formulated to avoid possible test bias and discrimination in placement procedures. Subjects consisted of the total population of 139 Black freshman students and a sample of 139 White freshman students entering a Southeastern regional university in the fall of 1970. Separate regression analyses were performed for Black, White and combined (total) groups on several sets of data. Regression analyses consisted of English GPA on English ACT scores, Math GPA on Math ACT scores, Psychology GPA on Social Studies ACT scores. Analyses were also performed for first semester GPA on Composite ACT scores, and second and fourth year GPA on Composite ACT scores. Based on Cleary’s (1968) definition of test bias, the results indicate that a single regression plane cannot be used to predict grades for Blacks and Whites, Current University placement guidelines were found to place Blacks in courses where their probabilities of success are lower than that of their White counterparts. It is recommended that a more flexible placement policy be instituted in order to avoid challenges of bias and/or discriminatory placement practices. It is recommended that individual students decide whether or not to enroll in a particular course. This decision is to be aided by updated University placement guidelines (based on regression equations) issued to faculty advisors, along with reference to updated expectancy tables.
Identifer | oai:union.ndltd.org:WKU/oai:digitalcommons.wku.edu:theses-2832 |
Date | 01 May 1976 |
Creators | Ungarean, Robert |
Publisher | TopSCHOLAR® |
Source Sets | Western Kentucky University Theses |
Detected Language | English |
Type | text |
Format | application/pdf |
Source | Masters Theses & Specialist Projects |
Page generated in 0.002 seconds