Testy over school tests

October 30, 2000

Before either George W. Bush or Al Gore becomes the "education president," they should take this simple test:

Which of the following is true:

1. All standardized tests are imperfect measures of student ability.

2. Not all tests can be compared reliably with one another.

3. Tests are being improved to become better indicators of achievement.

4. Tests are necessary to judge the progress of students and schools.

5. All of the above.

If both presidential candidates chose "5," there might be more light and less heat in the latest political dust-up over educational testing.

The heat comes from a study by the Rand Corporation, a California think tank that's normally neutral enough and smart enough not to release a politically charged report just days before an election.

The bad timing alone casts doubt on the study's integrity. But critics are also taking issue with its conclusion that students in Texas did not do as well on the National Assessment of Educational Progress (NAEP) exam as they did on their own statewide exams. Comparing the two tests doesn't make sense, they say. And anyway, Texas students generally fare very well compared with those in other states.

Nonetheless, Mr. Gore has used the study in TV ads to undercut the Texas governor's educational claims.

It's doubtful such charges will sway voters in deciding which man will do the best job of improving local schools from Washington. The candidates are better judged on their clear differences over federal education policy.

Still, a national debate on school testing has its merits, even if just to educate the candidates.

The statewide standardized tests given in Texas may indeed fall short on rigor and scope. The same, however, could be said of tests used by other states. This is hardly news.

The Rand report, using data from 1994 to 1998, found little improvement in Texas on the NAEP tests, unlike the state tests. But that conclusion must be weighed against the fact that the state tests are based on school curriculum in Texas. Students and teachers put more effort into preparing for them because the results have direct consequences.

Last July, a much lengthier Rand study found significant progress in Texas on the NAEP math tests. That study quickly became exhibit No. 1 for the Bush camp. It showed that black and Hispanic students in Texas were advancing compared with similar students in other states. And it went to some pains to adjust its findings according to socioeconomic factors, like the high number of non-English-speaking children in Texas.

Do these two pieces of research cancel each other out? No, they merely show that researchers can study similar data and draw different conclusions. The school-accountability program under way in Texas - which incidentally began under Bush's Democratic predecessors - still has proven merit.

This battle of the studies shows the complexity of assessing education progress. It should cause both candidates to be a bit humbler in asserting what they can do from the White House to reform education.

(c) Copyright 2000. The Christian Science Publishing Society