I occasionally teach test prep courses and have been interested in standardized tests for a long time, so a recent Atlanta Journal-Constitution news article on how closely Georgia’s SAT scores align with family income piqued my interest.
Critics of the two main college admissions tests — the SAT and ACT — complain that the tests are narrow measures of students’ abilities that unfairly reward students from the highest socioeconomic classes that can afford test prep. They point out that the tests are highly correlated with family socioeconomic status. So, are the tests necessary?
The answer is yes, for at least two reasons. First, the SAT/ACT is often the only objective measure by which students from different school systems may be compared. Second, the SAT/ACT do a good job of predicting college success, particularly when combined with high school grade-point average and other standardized tests such as SAT II’s and AP exams.
Controversy surrounding the tests usually centers around the tests’ validity, specifically their predictive validity. A test is valid if it measures what it is supposed to measure or predicts what it is meant to predict.
The SAT and ACT are intended to predict freshman grades in college and more generally, college success. The College Board, ACT, and independent researchers have conducted hundreds of validity studies regarding the tests.
Synthesizing those studies, Wayne Camara, vice president of research and analysis at the College Board, wrote that students’ performance on the SAT is related to “freshman and cumulative college GPA” as well as to “the need for remediation when entering college” and “persistence and graduation.”
A 2006 report commissioned by the College Board and conducted by independent investigators analyzed whether a high school GPA or the SAT was a better tool with which to predict college success. The researchers found that the SAT was “as good as or better than high school GPA in predicting high levels of college success.”
What’s more, the National Research Council of the National Academy of Sciences reported that SAT scores predict performance as accurately for blacks as they do for majority applicants.
ACT’s research found that “if an institution wants its admission criteria to reflect collegiate academic proficiency or ultimate level of degree attainment, ACT scores should carry greater weight than high school grades.”
A 2012 study in the journal Psychological Science concluded that the SAT’s ability to predict a student’s future college success is strong, especially when used in conjunction with high school grades, and that the use of the SAT has equal predictive accuracy for students from across a range of socioeconomic groups. According to that research, “it is not the case that the SAT is nothing more than a proxy for socioeconomic status.”
What critics of the SAT/ACT often ignore is that the correlation between the tests and college success is artificially weakened because the students selected at various colleges have a fairly narrow range of scores. That is, the students admitted to Georgia Tech, for example, will generally have scores bunched together at the high end, and a 2300 SAT — the highest possible score is 2400 — is not going to be a much better predictor than a 2200.
Were elite colleges to admit students they are already accepting and students now attending community colleges, there would be a considerably stronger correlation between SAT scores and college grades at those schools.
As long as the SAT and ACT remain an essential part of college admissions, they are likely to have their detractors.
However, unless and until we come up with a better means for objectively differentiating applicants to our nation’s top universities, we need those standardized college admissions tests.
Patrick Mattimore is a former high school teacher and an adjunct professor in the Temple University/Tsinghusa University LLM program.
