In today's Hoya (the Georgetown University paper), we find yet another well-meaning, yet naive, article about efforts to close the SAT score gap between subgroups, most particularly whites and minorities:
...Recent statistics show that the minority SAT gap persists: minority students generally score 150-200 points lower on average than their white counterpart, according to the College Board. Statistics released by District of Columbia public schools indicate that the gap is particularly prevalent in the Washington, D.C. area. In 2003, black students’ average combined scores were 438 points below that of whites’, and Hispanics’ average scores were 446 points lower than white students’...
Economic factors may also help explain the gap. Students from households making under $10,000 a year post combined scores of 864 on average, while students from households making $100,000 and up post scores of 1123 on average — a difference of 259 points. In a purely statistical sense, students’ SAT scores increase by one point for every $350 their parents take home.
Dennis Williams, the director of Georgetown’s Center for Minority Education Affairs, criticizes the SAT, emphasizing “the scores measure family income more precisely than they measure how well you can do in college.”
No, the scores measure how well you are prepared for college, which is related to how much money your parents have, because more money often means better schools, more tutoring - and more educated parents to start with. Ms. Williams needs a refresher course on correlation coefficients, and how a correlation between X and Y may appear because both are correlated with Z.
And even if SAT scores were correlated just with income per se, and not with the opportunities for educational accomplishment that extra dollars afford, does Ms. Williams provide any reason for us to assume that students with low SAT scores/low incomes will do well when placed in college? No, she does not, but SAT critics rarely provide any of this type of supportive reasoning.
Parental involvement also plays a large role, Williams says. Parents who did not themselves go to college — while not less involved in the lives of their children — are less likely to understand the college admissions process. Educated parents are in a position to help their children more, says Williams.
Again, what about this is supposed to convince me that the SAT should be removed, so that kids who have not accomplished as much, be it due to parental involvement or income, will have a greater shot at college? College isn't a babysitting service that should be open to those with the most need, as opposed to those with the most potential.
Another argument involves some of the SAT questions themselves. Although Educational Testing Services, which produces the test, claims it is trying to eliminate any bias inherent in the exam, Jay Rosner disagrees...“The test company uses a completely neutral, colorblind system for picking questions. However, that system predictably, consistently and reliably yields questions that favor whites dramatically over other subgroups,” Rosner says. The result of these rules is that the only questions that end up being on the SAT are “white-preference questions,” he says, which are questions prescreened and chosen to be on the SAT that white students answered correctly more often than other students.
You can go here to read my take on this blather, or you can simply ask yourself a few simple questions that go far towards demolishing his arguments. Do you really believe that all 2000 or so of the ETS and College Board employees are devoted to making sure black students cannot get SAT items right? For what reason would they do this, other than that they believe only white people should be going to college? Do you really think the KKK is running ETS? And if you believe that ETS and the College Board have the power and will to construct the test in this way, why do Asians outscore whites on the math portion of the exam? Don't you think the KKK would control for that, too? What about that "Other" group that's doing so well? Is that just because ETS can't purposely create items that are biased by race if examinees don't tell ETS what their race is?
You cannot go far with the likes of Mssrs. Rosner and Freedle without having to accept as fact the implied statement that ETS and the College Board are engaged in a deliberate conspiracy to foil black college hopefuls. If you really believe that, I know more than a few psychometricians who would be happy to help you readjust your tinfoil hat.
While he acknowledges that part of the discrepancy is in test preparation, Rosner asserts that if they let him pick the questions, “there would be a dramatic close in the gap.”
And if you let me tinker with your bathroom scale, there will be a dramatic change in your BMI measurement - but you won't be any skinnier. Yet another writer has missed the chance to point out that closing the gap at the end does nothing of value for minority students, because they're really suffering from an achievement gap, not a score gap.
Minorities may also face difficulty in preparing for the SAT. How well a student performs seems to be related to how familiar they are with the test; both Kaplan Inc. and The Princeton Review Inc. preach that the test is teachable. Their courses, however, are often priced close to $1,000 — a hefty fee for students in low- and middle-income families.
And this reflects badly on the SAT...how? The College Board didn't set those prices. Another way to look at this is to say that Kaplan and the Princeton Review are profiteering from the struggles of minority children trapped in bad schools. Test prep is virtually free for any child motivated enough to look for it.
Of course, these programs are garnering huzzahs for reducing their prices for minority students, so I suppose no reporter would dream of taking these companies to task for setting their prices so high to begin with. It's much easier for everyone just to blame the test.
What follows is some interesting information about foundations and efforts to increase minority performance on the test:
In Washington, D.C., specifically, the Hoop Dreams Scholarship foundation raises money to help students succeed through a series of three-on-three basketball tournaments. The Princeton Review provides courses for underprivileged students in conjunction with this program free of charge. A program called “Keys for Life” is also run through a partnership between The Princeton Review and the Child and Family Services of D.C.
In the District, Kaplan has started the Good Sports program, a more holistic approach to the college program. High school athletes can apply to this program, which is comprised of a Kaplan SAT course and college counseling and is conducted by Dr. Kpakpundu Ezeze, director of Future Quest Inc., an educational consulting firm. Both parts of the program are provided free of charge.
Ezeze says that the program is looking for students who are “not necessarily high SAT scorers. They can be average. We’ve had students that have had combined 700s, 800s, 900s.” The only requirement of the program is that the students play a sport in high school and that they are willing to put in the time over their junior and senior years.
Through the Good Sports program, Margaret Arbuthnot raised her SAT scores to a 1420 and is now a freshman at Princeton. She says she is grateful for the services the program provided but that she was motivated to succeed on her own. “If I had to pay a lot for it, I wouldn’t have done it,” she says. “I didn’t know how important SAT scores were in admissions because I’m the first kid in my family to go to college. I recommend test prep, but I think it’s outrageous that people have to pay a thousand dollars for it.”
The author, Lindsay Breedlove, also points out a recent College Board study showing that retaking the exam (which is cheaper than a prep course) tends to help students prepare better, and raise scores higher, than do the expensive prep courses. Unfortunately, she then follows up with the Ridiculous Professor Berube and his fantastically idiotic golf-handicapping suggestion, which is presented here with a completely straight face.
So, Rosner, Freedle, and Berube, along with other testing critics, are given space in this article without so much as a single quote from a College Board employee in response. This sort of imbalance exposes the ideology/gullibility of the writer and keeps this article, which does convey some good information about scholarship programs, from being entirely believable.
Update: Ms. Breedlove should take a look at this editorial that urges readers to take minority test results "seriously;" this means skipping over the pat answers about score differences and getting to the heart of fixing the educational issues:
Most teachers will tell you students learn in different ways and at different rates. To look for common indicators may be appropriate in view of the national learning gap, but they also run the risk of degenerating into stereotypes that do more to excuse poor performance than to correct it.
The data itself [subgroup data on the MEAP] is a new benchmark. The state released it in accordance with the federal No Child Left Behind law. It requires school districts to track math and reading or language scores by race. That should further ensure all students receive the best education available.
The measure also should encourage educators to look critically at teaching methods that seem to work and the ones that don't. Cleveland Elementary School, which has a large minority enrollment, saw black fourth-graders outscore their white classmates in math and language arts. In the Anchor Bay school district, black students performed better than white students in math.
Bridging the gap between white and minority students must be an essential goal of quality education. It must be approached critically and objectively.
And not with loopy conspiracy theories or inane athletic analogies.
Posted by kswygert at October 21, 2003 11:46 AM