Recently, I made the case that the “merit plea” opposing affirmative action is an inherently racist argument. In the blog, I briefly mentioned that the SAT is racially biased and not an adequate measurement of merit, as many would lead us to believe. The test is and has long been biased against African-Americans.
For me, grossly and consistently unequal outcomes almost always prove racist intent. The annual outcome of Whites out-scoring Blacks on the SAT demonstrates intent, proves the test is biased. I do not need evidence that shows racist intent. The outcome, for me, is enough.
However, this evidence would be nice to present to the legions of Americans who worship the mythology of neutrality of the SAT. To my satisfaction, after reading the “merit plea” blog, a Diverse reader pointed me toward a study in the recently published book, SAT WARS: The Case for Test Optional College Admissions.
I want to disclose to you the findings in Jay Rosner’s chapter, entitled, “The SAT: Quantifying the Unfairness Behind the Bubbles.” The startling new evidence of racist intent by the SAT test makers is mind-boggling, as my thoughts are still reverberating from its nasty implications.
Each year, when students take the SAT, one of the sections on the test is not scored. Instead, the Educational Testing Service (ETS) “pre-tests” the questions in the unscored section for potential use on a future SAT. During this pre-testing process, test developers also gather the race and gender of the test taker.
If questions perform well, then they are used on a scored section in the future. If they perform poorly, then they are scrapped. How does the ETS judge the performance of a question?
“Each individual SAT question ETS chooses is required to parallel the outcomes of the test overall,” writes Rosner of The Princeton Review Foundation. “So, if high-scoring test-takers — who are more likely to be White (and male, and wealthy) — tend to answer the question correctly in pretesting, it’s a worthy SAT question; if not, it’s thrown out. Race and ethnicity are not considered explicitly, but racially disparate scores drive question selection, which in turn reproduces racially disparate test results in an internally reinforcing cycle.”
This discriminatory process is known as “point bi-serial correlation.” According to Rosner, it is “a key methodology used by psychometricians to construct admission tests such as the SAT, ACT, GRE, LSAT, GMAT, MCAT and many other bubble tests.”
Rosner analyzed a data set of 276 verbal and math questions from the 1998 and 2000 SATs. He found what he calls “Black questions,” in which more Blacks than Whites answered correctly in the pre-testing phase.
“But it appears that none ever make it onto a scored section of the SAT,” Rosner says. “Black students may encounter Black questions, but only on unscored sections of the SAT.”
Instead, SATs only contain what he calls “White questions,” in which more Whites than Blacks answer correctly in the pre-testing phase. That is the principal reason why Whites consistently perform better. They are supposed to. The questions are literally geared toward them, as test developers are mandated to recreate the norm, the norm of White males outperforming their peers.
We no longer need to use unequal outcomes to deductively prove racist intent in the SAT. The use of point bi-serial correlation, the eliminating of the Black questions, the retaining of White questions is the heartbeat of racism in the SAT.
It is subtle, not loud. The rhythm of discriminatory test-making is regular. And it will continue to clandestinely pump out inequality unless we do something about it.
Dr. Ibram H. Rogers is an assistant professor of Africana Studies at the University at Albany — SUNY. He is the author of The Black Campus Movement: Black Students and the Racial Reconstitution of Higher Education, 1965-1972 (Palgrave Macmillan, 2012).