 |
  |
 |
 |
|
|
 |
 |
 |
 |
Opinion from the December 17, 2003
Reveal
the racial data on SAT scores
By Mark Franek
PHILADELPHIA — This month, nearly a million high school seniors
will put the finishing touches on their college applications. They will
write stellar essays, garner recommendations from favorite teachers, and
purchase new printer cartridges.
Pity, most of their efforts are purely cosmetic. From the colleges' standpoint, what matters most is something still not even a part of the application: the SAT score.
For many students, the SAT experience is not unlike an arranged marriage, one that forces students into years of counseling and prep before the big day. Each new generation of seniors would prefer a divorce from the grueling test that fails to reflect their entire high school experience, not to mention important traits like honesty and perseverance. Instead, they get coached. It should be no surprise that the College Board, which owns and administers the SAT, would like to see this very profitable relationship continue indefinitely.
The response to an essay on the SAT's alleged racial bias, which appeared in the Harvard Educational Review last spring, is a case in point. The author, Roy Freedle, retired in 1998 from the Educational Testing Service (which develops the test for the Board) after 31 years. He is not a disgruntled ex-employee. Mr. Freedle suggests that the SAT could be revised to include a score that would benefit African-Americans, who seem to do relatively better, on average, on harder questions. This supplemental score might help colleges make more informed choices, particularly when dealing with minority applicants from disadvantaged backgrounds.
The board's response to Freedle was a one-page unsigned article on their website (www.collegeboard.com). The article called Mr. Freedle's study flawed, mired in technical problems, and easily explained by students who merely guessed.
The board's disregard for any type of challenge is suspicious. It seems
to believe that the test, and the more than 1.4 million teens who take
it annually, exists in some kind of vacuum where all cultural differences
can be kept constant and all variations in scores explained by chance.
But any high school teacher worth his or her shiny apple will tell you:
No test is completely bias-free, and not all tests are equally fair. That's
why the trend at every high school in the country is to include more assessments
during the course of a year — not fewer — to test and honor
multiple intelligences. The real question is not whether the SAT contains
culturally biased questions, but whether the bias is systematic, persistent,
and statistically significant.
Instead of discrediting its critics, the Board should be willing to organize
all of its data into categories of race and socioeconomic background and
show how each group scores on the medium, easy, and hard portions of the
test. Furthermore, information about the race and gender of their current
test-writers should be disclosed.
Next, the board should release the racial breakdown of the exponentially increasing number of test-takers who are granted testing accommodations on the SAT. (As of this year, the Board is no longer reporting scores to colleges with any kind of footnote indicating that the test was taken under nonstandard testing conditions such as time-and-a-half, or double-time for students with learning differences.)
This data will show that whites — particularly higher-income whites — receive testing accommodations on the SAT more than any other group. Why? Because they can afford expensive testing evaluations. It should be no surprise that well-educated parents with means are better able to navigate the board's requirements at their children's high school in order to benefit from testing accommodations.
Full disclosure of this data is not a national security risk. It will, however, raise awareness of what areas in administering the SAT need attention or systematic restructuring.
Finally, the board should stop denying that coaching and expensive courses either don't help or provide only modest gains to a student's score. Give us teachers — and the public — a break. There is a reason the board provides links on its website to order its own practice tests and test-taking tips in book form or on CD-ROM ($395 for the school edition).
The board is currently working on the new SAT, due out in March 2005. It knows — though it doesn't want to admit it publicly — that its Goliath has spawned a multimillion-dollar "beat the test" industry that doesn't help disadvantaged kids, particularly those in the inner city. Those kids could tell the Board a thing or two about survival in the concrete jungle. But they sure won't be tested on it.
Mark Franek is the dean of students at the William Penn Charter School
in Philadelphia.
