Printer-friendly versionTaking the Scholastic Aptitude Test has become a rite of passage for millions of American school kids. The path to the middle class lies through getting a college education. And getting into college—especially at a prestigious school—requires a good score on the SAT.
Psychologist Roy Freedle has argued since 1987 that the test is biased culturally against blacks and Latinos. The Educational Testing Service, the people who administer the SAT, have dismissed his work. However, a paper recently published by the Harvard Educational Review will make Freedle’s work harder to ignore. According to the Review:
“By replicating Freedle’s methodology with a more recent SAT dataset and by addressing some of the technical criticisms from ETS, [the authors] confirm that SAT items do function differently for the African-American and white subgroups in the verbal test and argue that the testing industry has an obligation to study this phenomenon.”
This blog post from The Washington Post can help explain the issue more. Also, Teaching Tolerance has explored the issue of culture and standardized testing here, here, and here.
Washington Monthly writer Daniel Luzer says that Freedle and others have suggested fixing the SAT by changing how it is worded. Luzer points out that this would superficially address the issue, but it does nothing to attack the underlying cause. “The trouble with this is that it isn’t actually the SAT scoring that hurts black students so much as it is poor education that hurts black students,” Luzer says. “Adjusting the score wouldn’t fix that problem.”
Education reformers obviously need to tackle both issues—the bias of this critical test and the criminal under-education of black and Latino students.
Comments
Not only is the SAT racially
Not only is the SAT racially and culturally biased, it is also biased with respect to students with learning disabilities. In other words, if one tests well, regardless of race or SES, one can access a wider array of post-secondary options for those colleges and universities which place a heavy premium on standardized test scores, especially those colleges and universities which use the SAT as a mechanism to weed out candidates who may otherwise be qualified.
There is a growing cadre of colleges and universities that are viewing the SAT in its proper light, i.e. it has no real validity in as far as predicting college success over a four or five year period. However, more colleges and universities have to be willing to use other markers of potential success when evaluating prospective candidates. Mount Holyoke College, for example, conducted a study several years ago to determine if making the SAT an option for admission would have a negative impact on the quality of its applicant pool. It did not.
Having said the aforementioned, the only aspect of the SAT that isn't biased is the math. At least, it wasn't when I took the math component of the SAT.
How is being biased against
How is being biased against those who don't test well a bad thing, given that its purpose is to qualify one for college entrance?
It's not like someone who "tests poorly" isn't going to have to take tests in college.
The reason for issue lies in
The reason for issue lies in the fact that poorly educated students will never do more than test poorly because - they have a poor education. It is helpful because it brings additional light to educational discrepancies, which those in power continue to ignore. In the same light, it shows that the SAT testing language privileges cultural norms that are actually only normal to a select group of people. Finally, it rightfully tears away at the assumption that standardized testing is mostly objective.
"In the same light, it shows
"In the same light, it shows that the SAT testing language privileges cultural norms that are actually only normal to a select group of people."
Given that the SAT is supposed to be a college entrance exam, if this is the language that one would expect in college, then it should privilege these cultural norms.
"Finally, it rightfully tears away at the assumption that standardized testing is mostly objective."
No, it doesn't
What bias is there? Dont you
What bias is there? Dont you think that maybe the results are what they are because blacks and latinos are on average worse educated? The reasons for this phenomenon are various, from cultural through genetic to discrimination. But test results have to reflect this, and you cannot try to artificialy inflate scores of some groups. That would make the test biased!
I am Latina and I think that
I am Latina and I think that racial bias in SAT tests is utter nonsense. Check out JunkScienceMom's blog and her article in response to the blog post from The Washington Post. It is called "Junk Reporting – the Hidden Truth Behind Claims Of Racial Bias In the SATs" http://bit.ly/aMthHF
JunkScienceMom cites a paper
JunkScienceMom cites a paper by University of Minnesota researchers Paul Sackett, Matthew Borneman and Brian Connelly:
"This assertion consists of two parts. The first is that mean scores are lower for certain minority groups. This has long been known to be true; we offer details below. The second is that these mean differences can be interpreted as evidence of bias in the tests; this inference is unequivocally rejected within mainstream psychology."
The cited paper and JSM fail to address Freedle's findings that black test takers actually outscore white test takers on questions that appear to be more difficult for all test takers, that is questions that are more likely to be answered correctly only by those who do very well on the test. White test takers did better than blacks only on questions that were likely to be answered correctly even by those who didn't do so well on the test. Freedle's findings cast doubt on the validity of the test since questions should discriminate among test takers based on how they performed on the entire test (or the set of skills the test purports to measure) not based on "irrelevant" characteristics such as race.
Freedle's proposed fix involves providing, in addition to overall scores, scores that exclude questions that were answered correctly even by those who did not do well on the entire test. Some seem to consider this "dumbing down" the test but it's hard to see how eliminating the easy questions from the score could do this. Questions such as these appear to discriminate on the basis of race rather than knowledge, skill, or prediction of college success. The fact that the SAT can predict college success is no more evidence for it's lack of bias than the racial difference in mean scores is evidence for bias.