ERIC Number: ED478832
Record Type: Non-Journal
Publication Date: 2003-Apr
Reference Count: N/A
Reconsidering the SAT-I for College Admissions: Analysis of Alternate Predictors of College Success.
Armstrong, William B.; Carty, Heidi M.
The University of California is engaged in the elimination of the Scholastic Assessment Test I (SAT-I) Verbal and Mathematics tests as a requirement for freshman admission. Opponents of the SAT-I argue that the tests do not measure the outcomes of the high school curriculum and hence do not reflect student learning in secondary school. Proponents counter that while the SAT-I tests are imperfect predictors, they perform a useful role in selecting applicants who have a strong likelihood of college success. This paper discusses the policy background of this debate and compares criterion-related validity evidence for the SAT-I and SAT-II tests. The study used data from applicant and enrolled student records at a large, highly selective research university for approximately 18,000 first-time freshmen. The findings suggest that although the SAT-II tests show stronger criterion-related validity than the SAT-I tests, the differences are modest. It is also found that the predictive validity of the SAT-I mathematics test improves for students from lower income levels and socioeconomic backgrounds. In addition, data from the third achievement or subject test of the SAT-II indicate a low correlation of scores in this test with freshman year grade point average (GPA) and high school GPA. This suggests that the third achievement test may be less of a curricular measure, or of less value, in accounting for variance in first-year performance. (Contains 7 tables and 11 references.) (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A