State: School tester hopes to ID 'strivers'

School tester hopes to ID 'strivers'

Some students do better than expected on tests such as the SAT, given their background. But critics say compensating for such factors is a bad idea.

By STEPHEN HEGARTY

After years of criticism that its test is biased against minorities, the poor and students from bad schools, the company that administers the SAT college admissions test is working on a way to identify students who overcome disadvantages that tend to pull scores down.

The new formula, which still is being studied by the Educational Testing Service, is designed to identify "strivers": students who score higher than expected given their background. The factors the formula would consider include family income, parent education, the quality of the student's high school and race.

ETS officials say it will be a couple of months before they are ready to release the study and draw conclusions. But the proposal already is drawing interest around the nation, coming at a time when race-based admissions are prohibited in some states and under attack in others.

"If this offers a framework to get a more accurate predictor of success, why wouldn't we be interested?" said Keith Goldschmidt, director of public information for the Florida Board of Regents. "There's no formal review of our procedures or anything, but some of our staff people are definitely looking at it."

And, as happens with any proposal that deals with race, issues of fairness and college admissions, the "strive" proposal is drawing controversy.

"All this does is legitimize the stereotypes of black kids and Latinos," said Ward Connerly, the California businessman who led the anti-affirmative action campaign in that state. "This is creating a profile, saying, "This is all we expect of you as a black kid.'

"I think there is great merit in trying to find out if some kid didn't perform well due to some disadvantage. But I don't know how you can quantify a person's background."

The strivers concept is of particular interest in Florida because Connerly is hoping to eliminate race as a factor in college admissions here, as he did in California and in Washington, through a statewide ballot amendment.

"The Ward Connerly initiative has caused everyone to pause and look at what affirmative action is and what role it plays in college admissions," Goldschmidt said.

Although the concept is controversial and the formula is complex, the basis of the idea is almost universally accepted: Some disadvantaged children have a lot more academic potential than their test scores would indicate, and they should be given a chance despite their mediocre scores. The question facing admissions officers is how to root through all the applications and find those diamonds in the rough. The SAT people hope to provide a tool to do just that.

The project is focused on students whose SAT scores fall into the range of 1,000 through 1,190. Researchers felt that students who score 1,200 or above already tend to fare well in admissions and in college. (The highest you can get on the SAT is 1,600.) Strivers are defined as students who score 200 points higher on the SAT than might be expected given their background.

Factors the formula would consider include the family's socioeconomic status, race and ethnicity, the quality of the high school, the student's grade point average and the academic rigor of the student's high school courses. Race could be added to or subtracted from the list of variables. That might be important in a state such as California, where it would be illegal to consider race in college admissions.

That kind of information could easily be collected; students who take the SAT commonly fill out what is called the Student Descriptive Questionnaire, which provides information about the student's demographics and schooling. ETS researchers have been able to test their "striver" formula by essentially looking backward: taking students who already finished college and looking back at their demographic profile coming out of high school. In that way they could pick out strivers and see how they did in college.

"When I first heard about it, I was kind of excited," said John Barnhill, director of admissions at Florida State University. "I thought it might make a tough job a little easier."

For this year's freshman class, Barnhill's office faced 19,000 applications. (Roughly 6,400 were accepted.) Barnhill said that students with impressive grade point averages (say a 3.5 or higher) or high SAT scores (say a 1,200 or higher) tend to be accepted. Others with low GPA's (say a 2.5 or lower), or low SAT scores (say 900 or lower) tend to be rejected.

"What you're left with are a lot of files in the middle," Barnhill said. "You try to look at as many factors as you can -- not just the test scores. That's why the strivers concept sounds promising. It gives us some other things to consider."

The strivers idea, which has been likened to a golfing handicap, could theoretically be used in other tests, especially tests where minority students and poor students tend to score lower. But don't expect Florida's premier test, the Florida Comprehensive Assessment Test, to delve into strive factors any time soon.

"We have always said we want high standards for all kids," said Karen Chandler, spokeswoman for Education Commissioner Tom Gallagher. "It seems (the strive concept) would be inconsistent with having the same high standards for everyone."

Connerly and other critics see the concept as well-meaning but deeply flawed.

"This new wrinkle is the ultimate confirmation of what we've been saying for years: There are people who come from difficult circumstances who don't do as well on this test, and that's because the test is flawed," said Jeff Rubenstein, senior director of research for the Princeton Review, the test preparation service. "This is a Band-Aid, intellectual gymnastics. What we should be doing is looking at other methods, something other than the test."

But admissions officers, especially, are open to any legitimate new methods for telling them more about their applicants.

"There could be some value in the idea of weighing certain factors," said Goldschmidt with the Florida Board of Regents. "We have recognized that students don't all come from the same neighborhoods, the same circumstances."