Jay & Valerie Debate Cheating
Class Struggle and colleague Valerie Strauss, of The Answer Sheet, disagree on what to do about accusations of D.C. educators covering up bad results on the annual DC-CAS tests. Jay argued for retests in his Monday column.
Valerie saidthat wouldn't work. Here they try to work it out.
JAY MATHEWS: I hear you didn't like my solution to the DC test cheating dispute. Where did I go wrong?
VALERIE STRAUSS: I wouldn't say you went wrong. It would be unwise for me to ever say that, Jay. I would just say I don't agree with you. You think the kids should retake a standardized test to prove they learned the material after concerns about cheating raised questions about the validity of the results. I think that the kids shouldn't retake the test because: a) the conditions under which the alleged cheating still remain, and b) the score on one standardized test doesn't mean much statistically. It doesn't tell us what it purports to tell us--that kids learned certain material and know it and understand it. So where am I wrong?
JAY: Explain to me why we couldn't change the conditions under which the alleged cheating occurred--have proctors at the retest especially appointed by DC's school board office. On average, one standardized test doesn't mean much statistically, but when you are looking at individual students, you can get a pretty good idea if the original test score was believable, or not.
VALERIE: I grant you that you could change the conditions of who proctors the test. But the second score may not tell you anything about the first score. The kinds of variables that could have resulted in a first test score that doesn't show a student's true ability could be present in the second score. I have tried things once, done poorly, and the second time done extremely well. And vice versa. The latter result wouldn't mean I had cheated.... I won't even mention the questionable validity of a student test that can affect how much a teacher earns (with the assumption being that the teacher was solely responsible for a high score or a low score.) Oh, I just did. Sorry.
JAY: As I said in that column, I am more interested in discovering whether these kids---possibly betrayed by cheating educators who covered up their flaws----have actually learned what they were taught. We are always going to have cheating. We have to look beyond that to see how it has affected kids' progress, in isolated cases like this where bad deeds may have occurred. How am I going to do that if you don't let me retest?
VALERIE: I agree that we have always had cheating and we always will. I also agree that it is important to discover whether the kids have learned what they were taught. My problem is that I don't believe that one standardized test can prove that the kids learned what they were taught. So retesting won't help any better than the first test. I am not anti-test. If we didn't have tests we would invent them. But the high-stakes standardized test variety don't tell us much of anything. That's my view, and I'm sticking to it.
JAY: Ah, well. Your argument makes sense, but it has the disadvantage of being the way we have always handled these situations. Retesting would be a relatively new approach. We might learn something from it we did not expect. But I fear there are no school administrators out there crazy enough to try it.
The comments to this entry are closed.