skip to main content skip to footer

Trends in the Predictive Validity of the Scholastic Aptitude Test SAT

Author(s):
Fincher, Cameron
Publication Year:
1990
Report Number:
RR-90-13
Source:
ETS Research Report
Document Type:
Report
Page Count:
82
Subject/Key Words:
College Board, Georgia, Predictive Validity, Scholastic Aptitude Test (SAT), Test Validity, Trend Analysis

Abstract

The objectives of this study are: 1) to summarize the changes in the SAT's predictive validity that have been observed in the University System of Georgia (USGA) and other similar institutions over the past two decades; and 2) to identify trends where they can be described with reasonable accuracy. Inquiry was made by mail or phone to over 200 colleges and universities and varying responses were received from 53 institutions. Longitudinal data were obtained from 14 institutions that supplied unusually good information for inclusion in the study. Four earlier studies provided baseline data for this study. The bulk of this report is presented in the graphs presented in the two appendices: 1) Five-year Averages in Validity Coefficients for High School Averages and SAT Scores; and 2) Institutional Profiles of Cooperating Institutions. Within the report itself, data is presented on: 1) trends in five-year averages; 2) changes over ten-year periods; 3) changes in variability; 4) variations in validity coefficients; 5) variability and predictive validity; and 6) profiles of cooperating institutions. Conclusions are drawn from the various graphs and tables depicting variations in validity coefficients and from efforts to analyze the observed variations in validity coefficients. It is further concluded that 1) the SATs are still valuable as additional predictors, with high school averages (HSA), of college freshman grades; and 2) that the variations in the SAT's validity coefficients can as readily be explained by changes in academic policies and standards as by changes in the predictive efficiency of the SATs. (JGL) (82pp.)

Read More