Misused Education Stats in the Star's WaPo Editorial


1 comment
It's OK that I don't agree with the Washington Post editorial in Friday's Star, which says we should keep testing students in reading and math every year. I'll even hold my nose and not complain when it says the testing regimen beginning with the No Child Left Behind law is "a civil rights achievement that must not be undone." But when the editorial misuses statistics to make the case that achievement of poor and minority students has gone up because of the annual NCLB testing—nope, I won't let them get away with that.

The editorial says the NAEP (National Assessment of Educational Progress) test results indicate that poor and minority students have increased their achievement since 2001. I kinda like the NAEP test, because it's low stakes—a sample group of students are given the test, so no one can teach to the test, and no one loses their job over the results—and it has been given consistently since the 1970s. To the extent that you can put faith in any standardized test as a measure of student achievement, this is the one to put faith in.

And it's true, the scores of black and Hispanic students have increased since NCLB  became law, and they've increased at a faster pace than the scores of white students, so the gap between white and minority student scores has shrunk, even though the gap is still pretty dramatic. But the problem with giving NCLB any credit for the increase is, the scores have been increasing since the tests were first given forty years ago at about the same rate. If you follow the jagged, up-and-down lines on a graph charting the scores, you find that the slope from when the first tests were given to 2001, when NCLB began, is close to the same as the slope from 2001 to the most recent tests in 2012. The only group whose scores have slowed their ascent significantly is white 17 year olds.

Here are the two conclusions the WaPo editorial could have reasonably drawn using the NAEP scores from their beginning in the 1970s to the present. (1) The average scores of all students have increased over the past forty years, blacks and Hispanic students faster than white students. That means the meme that our schools are failures, that they have gotten worse over recent decades, is simply wrong. By the only consistent measure we have, student achievement has improved, minority students faster than white students. (2) The yearly high stakes testing since the beginning of NCLB hasn't increased the pace of improvement. At best, it's a wash—though, actually, the scores over the past four to six years have flat lined, which could mean that we're beginning to see the negative consequences of our obsession with teaching to the test.

The pro-test crowd says we need to keep on testing at the same obsessive pace because of all the valuable data the tests produce. Well, they should be more careful with the way they use their beloved data. If proving a point is more important than accuracy, they really don't need the students to take all those tests to generate all that questionable data. They can just make things up, like they always have. 


Showing 1-1 of 1


Add a comment