Cheating by teachers cannot: On Monday, we mentioned three key facts good mainstream journalists mustn’t discuss.
In the featured editorial of that day’s New York Times, one of those facts got deep-sixed.
In its featured editorial, the Times recommended the use of student test scores in teacher evaluations. Early on, the paper listed three bad things which can happen if student test scores aren’t used:
NEW YORK TIMES EDITORIAL (9/17/12): Traditional teacher evaluations often consist of cursory classroom visits by principals who declare nearly every teacher good, or at least competent, even in failing schools where few if any children meet basic educational standards.In our view, one of those scenarios makes good sense. The other two, possibly not.
As a result of this system, bad things can happen. High-performing teachers who have an enormous impact on student achievement go unidentified, and they often leave the district. Promising, but struggling, young teachers never get the help they need to master the job. And disastrous teachers who have no feel for the profession continue as long as they wish, hurting young lives along the way.
(Presumably, horrible test scores would be a strong sign that a teacher has turned "disastrous.")
Whatever! In principle, we aren’t opposed to the use of test scores in teacher evaluations. But there is a large problem with this idea—and in a lengthy editorial, the Times never brought it up.
That problem is cheating by teachers! As everybody knows by now, teachers and principals sometimes cheat when they administer high-stakes test. The higher the stakes become, the greater the incentive.
There are ways to address this problem, of course—though solutions may cost money. (The use of proctors, for instance.) But in a lengthy editorial about the use of student test scores, the Times never mentioned this existence of this obvious problem.
Teachers and principals cheat on tests! By now, even journalists know this is true—but mentioning this obvious fact is largely verboten. Quite routinely, mainstream pundits recommend the use of test scores in teacher evaluations without showing the slightest sign that they know this problem exists.
You can't discuss cheating by teachers! In the Times editorial, the silence was especially striking because the editors chose to write this:
NEW YORK TIMES EDITORIAL: Reasonable school officials understand that test scores, while important, do not reflect the sum total of what good teachers provide for their students. In Washington, D.C., where the evaluation system is now in its fourth year, school officials have decided to change the weighting of tests. Originally, value-added scores accounted for 50 percent of teacher evaluations; that has been reduced to 35 percent, with an additional 15 percent consisting of other goals (like the students’ mastery of certain skills) collaboratively arrived at by teacher and principal.Impressive! Unfortunately, Washington is one of the big school systems where massive cheating has occurred. But so what? The editors blew right past that fact as they recommended putting even more pressure on tests.
Every good journalist knows the rules: You’re not allowed to tell the world about those gains on NAEP scores. And you’re strongly encouraged to avoid all discussion of cheating by teachers. As a result, pseudo-discussions often take place, like this discussion on Monday.
The editors seemed to know—you’re not allowed to discuss teacher cheating. And good lord! Right next to their editorial, this letters column appeared.
Five letters appeared beneath this headline: “What to Do About Student Cheating.”
Students cheat—but teachers do not! No, we’re not making this up.
David Brooks turns the logic around: Last Friday, David Brooks advocated using test scores in teacher evaluations.
Our pundits rarely know much about schools. We thought Brooks’ logic was a bit upside down:
BROOKS (9/14/12): Though the final details are still uncertain, there will also be a serious teacher evaluation process [in Chicago]. The various elements of those evaluations will change for each teacher year by year, but, as teachers progress in their careers, student performance will become more and more important. That’s vital because various studies have shown that evaluations that rely in part on test scores really do identify the best teachers. Teachers who score well on these evaluations really do produce measurable improvements in their students’ performance for years to come. Rigorous teacher evaluations will give reformers a profound measuring tool.Citing unlinked studies, Brooks said the use of student test scores helps “identify the best teachers.” But when student test scores are used that way, it does increase a teacher’s incentive to cheat.
Brooks didn’t mention this problem. Surely, he knows it exists.
Brooks said the use of test scores can help us find our best teachers. Absent some solution to the cheating problem, student test scores seem better suited to help us find our worst (most “disastrous”) teachers. Presumably, no one cheats to produce bad scores.
If a teacher’s scores are disastrously bad, those horrible scores can almost surely be trusted.