UC: Some SAT tests better than others at predicting successes

The Associated Press
Friday October 26, 2001

OAKLAND — Subject-oriented SAT II tests do a better job of showing how students will do in college than the better-known SAT I college entry exam, according to a University of California study released Thursday. 

The study comes as UC faculty are considering a request by UC President Richard C. Atkinson that students no longer be required to take the SAT I. 

UC has required students to take both the SAT I (or ACT) and SAT II tests since 1968. 

The new study looked at the relationship between test scores and the records of nearly 78,000 freshmen from fall 1996 through fall 1999. 

The study found that SAT II scores gave a better indication than SAT I scores of how well students would do in college. Adding SAT II scores to high school grades gave a very good idea of student performance. Adding SAT I scores on top of that resulted in only a very slight increase in the ability to predict student success over the four-year period. 

That isn’t enough to justify the effort and expense of the SAT I, the study’s authors say. 

At the College Board, Amy Schmidt, director of higher education research, said she doubted the UC findings could be translated to students across the country. 

She said the SAT I and II, both owned by the board, are “both excellent tests. I just don’t think ... that most institutions would be happy with that (switching to the SAT II only.) Most institutions seem to really prefer the SAT I.” 

The SAT I is a test of language and math skills, based on how well students can reason. The SAT II is more content based, with each test devoted to specific subjects, such as history. Both tests are multiple choice. 

One of the criticisms leveled against the SAT is that it is culturally biased and unfair to disadvantaged students. Defenders say the test is fair; the problem is that not all students have the same educational opportunities. 

Robert Schaeffer of FairTest, which advocates less emphasis on standardized tests, called the UC study the “first truly comprehensive study showing how poor the SAT I is as a predictor.” 

Schaeffer, however, said the answer is not to switch to SAT II, but to stop using it altogether. 

Atkinson has proposed using SAT II tests as an interim measure while new tests are developed that would be more closely linked to the California high school curriculum. 

His proposal to drop the SAT I is being reviewed by the Academic Senate. 


On the Net: UC study, www.ucop.edu/sas/research/researchandplanning/