American children know a lot. And they can think for themselves. Researchers and politicians are fond of saying the opposite -- that students are appallingly ignorant. They want an agenda of "national goals," more tests and less chance for independent thinking, so they issue reports claiming massive failure.

But if you look behind their summary figures, the problems shrink like popped balloons. For example in a national writing test that drew considerable attention earlier this year, one question required 11th-graders to adopt a point of view about whether funding for the space program should be reduced and to write a letter to their senator explaining their position. They were given 7 1/2 minutes to do this. Only 3 percent did not answer; 18 percent answered without choosing sides. (With only 7 1/2 minutes, it seems reasonable to be noncommittal, unless one already has a position.) The other 79 percent of students explained their positions.

Is that "dreadfully inadequate" as Education Secretary Lauro Cavazos called it? The people scoring the test say many students lack reasons in their letters to senators. But the example they publish of a letter allegedly lacking reasons actually contains two: the need to solve other problems first and the need to benefit everyone, not just those involved in space.

Another question asked 11th-graders to read a "passage about frontier life and then to explain why modern-day food differs from frontier food." Only 2 percent did not answer in the 15 minutes allotted; 9 percent wrote about food, but gave no comparison. The other nine-tenths mentioned differences and discussed such conveniences as supermarkets, refrigerators, microwaves and junk food.

Is this "dreadfully inadequate"? The testers (Educational Testing Service, if you were wondering, for the National Assessment of Educational Progress) also marked grammar and spelling on these tests and found few mistakes.

A third question asked students to describe "a desirable summer job and to summarize their previous experiences or qualifications for it" in 7 1/2 minutes. This is a good task, since students should have already thought about this. Only 1 percent did not answer; 13 percent wrote vaguely with no details; 18 percent described a job but no experience. The other two-thirds gave answers considered adequate by the testers. Students need more frequent practice with this kind of writing.

In math tests you might think the right and wrong answers are clearer. Not so. Students were asked to apply data about the brakes in one kind of car to another, possibly different kind of car. Students who said this approach was unreliable were marked wrong. The square root of 17 can be positive or negative, but students who remembered the negative root were marked wrong.

A question about principal and interest asked how much money was "repaid." Interest is paid, not repaid, but that thinking was marked wrong. All seven problems published as examples of the most difficult level of math have serious weaknesses, and this is the only level that 11th-graders get wrong. The test also had severe time pressures: 40 to 60 questions in 15 minutes.

A recent Times-Mirror study claims dramatic drops in young people's interest in the news. They compare the number of people in the 1940s and '50s who had "heard or read about" a particular story with the number of people today who follow a particular story "very closely." Of course, the number drops. How could it be otherwise?

A geography test compares how many countries Americans and Europeans can find on a map. This was the 60th question out of 80 here, but the sixth of seven abroad.

It's traditional for each generation to think that the youngsters are going to the dogs; the new wrinkle is that we now try to prove it scientifically. But in spite of all complaints, U.S. workers still produce twice as much per hour as the Japanese. Usually people criticize our slower productivity growth, but that enormous lead helps explain why we don't grow as fast.

We have specific problems in some schools, some classrooms and some topics, like job applications and percents. We should look for specific issues such as these and solve them. But the myth of massive failure causes us to waste time and money on experts, tests and the creation of academic boot camps rather than devoting our energies to the specific improvements that are needed.

The writer is an independent education researcher and survey designer in Washington.