Grading on a curve

Common Cause Magazine, Summer, 1994 by Deborah Lutterbeck

If it were framed as a word problem it would go something like this: If 75 percent of the students in State X, which spends M+1 to educate its students, take the same test as 5 percent of the most select students in State Y, which spends M - 1, which state will have the higher test scores?

The answer seems obvious: Average scores go down as the number of test takers goes up. In real life, however, solutions are not so clear -- particularly when politics is thrown into the equation.

Take, for example, State Z, also known as New Jersey: The Garden State spends about $10,219 per pupil per year -- more than any other state -- but it ranks in the bottom half on the nation's standardized testing scoreboard. This apparent contradiction has given much fodder to critics of education spending such as former Education Secretary William Bennett.

His Empower America, a conservative Washington-based policy group, along with the American Legislative Exchange Council recently issued a report showing that while education costs rose almost 50 percent from 1972 to 1992 nationwide, scores on the Scholastic Aptitude Test (SAT) fell some 35 points. Proof positive, the report said, that school systems like New Jersey's, with its high costs and low test scores, are making a bad investment.

But it seems such critics could use a refresher course on some of the very skills the SAT attempts to measure: problem solving and reasoning. Lower SAT scores could be caused in part by a dramatic increase in the number of students who take the test. In fact, some say New Jersey ranks 39th on the SAT roster in large part because 76 percent of its students take the test.

Utah, which spent only $3,092 per pupil in 1992, ranked fourth on the SAT barometer. What gives? "About the only kids [in Utah] that are taking the SATs are those kids who want to go to Ivy League schools or are somehow competing to become National Merit Scholars," says David Nelson, the state's director of evaluation and assessment. Most college-bound Utah students take the American College Testing exam (ACT), which is required by state-supported colleges. Some 29 states favor the ACT, which means their SAT scores reflect results from a select pool of students.

This misinterpretation of SAT scores is perpetuated by the media and leads to misuse of the test by politicians, says Jay Comras, special projects consultant for the National Association of Secondary School Principals.

In some cases, state policy makers find themselves in an educational catch-22, says Robert Chamberlin, senior vice president of the investment firm Dean Witter. As states spend more on elementary and secondary schools, Chamberlin says, more students consider higher education, which results in more of them taking standardized tests, thus lowering overall scores. The lower test scores then may be manipulated by opponents of government spending (or proponents of privatization plans such as vouchers) to argue that increased state spending on education does not produce better-educated students.

There could be further complications in 1995, when the College Board changes the way it ranks the test scores. "The public is going to think that schools have gotten so much better," Comras says, when in fact scores will only seem higher. That's just one more reason public-school critics should take a closer look at the numbers they use to build their arguments, education experts say.