You are not currently logged in.
Access JSTOR through your library or other institution:
An Analysis of Types of Errors in the Selection of Minority College Students
Roy D. Goldman and Mel H. Widawski
Journal of Educational Measurement
Vol. 13, No. 3 (Autumn, 1976), pp. 185-200
Published by: National Council on Measurement in Education
Stable URL: http://www.jstor.org/stable/1433733
Page Count: 16
You can always find the topics here!Topics: Standardized tests, College students, Minority groups, Minority group students, Colleges, Universities, High school students, College graduates, Psychometrics, Matrices
Were these topics helpful?See somethings inaccurate? Let us know!
Select the topics that are inaccurate.
Preview not available
This investigation projected the selection errors which would result from the use of the SAT to select Black and Mexican-American college students. By using the regressions of grade point average (GPA) on high school grade point average (HSGPA) and HSGPA plus Scholastic Aptitude Test (SAT) scores, it was possible to project accept-reject/succeed-fail matrices for Black and Mexican-American students at four large universities. Overall, these groups scored considerably lower on the SAT than on either HSGPA or GPA. The addition of SAT did not improve, or only slightly improved, the validity of HSGPA as a predictor of GPA. Thus, the use of the SAT produced only minimal improvements in selection accuracy. However, these minimal improvements were achieved by decreasing false-positive selection errors while increasing false-negative ones. The net result was the admissibility of fewer minority students. The necessity of evaluating the costs of the different selection outcomes was discussed along with a possible explanation for weak GPA prediction. The problem of restriction of range as a factor in the attenuation of SAT validity was also discussed.
Journal of Educational Measurement © 1976 National Council on Measurement in Education