Happy Birthday?

Last week, "No Child Left Behind" turned six years old, a ripe old age by the standards of education reform waves. Old enough so we ought to be able to tell whether it's working. NCLB was going to make plain the flaws of American schools and apply the necessary pressure to make them shape up. One out of every four schools was unable to meet "adequate yearly progress" (AYP) last year, and one in seven was in "improvement" or "corrective action," which makes them subject to punishment if they receive federal Title I funds.

So is all that shaming and punishing helping kids?

Many teachers across the country complain that the law has undercut the teaching of important but untested subjects, like art and history. And many teachers hate the focus on test scores. On the other hand, some teachers do say that the pressure has focused their schools more clearly on educating children who have been low achievers.

But we also have hard, quantitative data to look at: If the measure of success is test scores, are the test scores rising?

For the answer, you can't just look at the state tests. Testing experts agree that when passing a particular test becomes very important, teachers tailor their teaching to that test and scores rise, without necessarily reflecting any real improvement that would help a child later in life. So we need to see increases in other tests that measure reading and math skills, not just the state tests. The National Assessment of Educational Progress (NAEP) fits the bill. NAEP is the only national achievement test. And researchers have done extensive research on how NAEP scores have changed.

Depending on the way you look at the data--which subjects, which subgroups, which ages--researchers have found different results: a little up, a little down. But nobody has found any major impact. No jump in the scores. No collapse, either. If you didn't know that a major change is going on in American education, you would never guess it from the NAEP scores.

There have been other periods which did see dramatic changes. During the 1970's and early 1980's, the achievement gap between African-American and White students was cut nearly in half. Even the most ardent pro-NCLB spin masters would have to agree that nothing like that is happening today.

Before NCLB took effect, there was already research on the effectiveness of high-stakes testing. Some states had high-stakes tests, others didn't. By comparing changes in scores in those two sets of states, researchers could see whether high-stakes tests raise scores.

Just like now, that earlier research turned up small declines here, small increases there, but found no major improvement. It may seem counterintuitive, but the evidence is strong that high-stakes testing does not make a whole lot of difference to test scores.

NCLB insists that teaching strategies must be backed by data. The term "scientifically based research" appears 111 times in the text. Doesn't it make sense to apply that idea to the law itself?

--Alain Jehlen