by David Safier
USA Today put together a research story about teachers and administrators cheating on high stakes tests, and the results, while perfectly predictable, aren't pretty. In the six states studied, including Arizona, the researchers found whole classes and whole grades at some schools where the students made unbelievable — literally, statistically unbelievable — progress. The next year, these same students' test scores plummeted.
And that's just in schools and classrooms where the researchers saw drastically improbable results. More cautious efforts to raise students' scores a few points will generally go undetected if no one blows the whistle.
How widespread is the cheating/tampering to raise students' scores?
An Arizona State University online survey of 3,085 Arizona teachers published last year found that more than half of the respondents reported that they had engaged in some form of cheating and knew colleagues who cheated. They recounted 19 ways to cheat.
The cheating ranged from erasing incorrect answers and filling in correct ones to telling students to redo answers and giving students extra time.
Fifty percent cheating sounds high to me. But clever ways to give student test scores a bump, some of them perfectly legit, some of them clearly cheating, are everywhere you have teachers and schools evaluated on test scores. When you mix test-as-diagnosis with test-as-evaluation, you're going to find neither result is accurate.
Here's an example of what researchers found in a Cincinnati elementary school, something which cropped up all too frequently. Third graders scored in the bottom fifth in the state in math. As fourth graders, they were in the top fifth. Then as fifth graders, they were back in the bottom fifth in the state. The district superintendent said it was all about great teaching in the 4th grade and poor teaching in the 5th.
The superintendent needs to put together a better story than that.
If the 4th grade teachers were truly great and there is such a thing as "value added teaching," those kids should have had some kind of halo effect and not slid 60 percentage points in a year. A 10 point slide maybe, 20 points maybe, but 60? That would mean the 5th graders forgot everything they learned the year before — highly unlikely.
And remember, teachers are increasingly being evaluated based on the value they add to their students' educations. Those 4th grade teachers should be in line for big bonuses based on their kids' phenomenal scores, and the 5th grade teachers would qualify as the Worst Teachers in the World.
If 4th grade teachers are "helping" their students raise their scores, the 5th grade teachers would be fools not to do the same. When some athletes dope to enhance performance, other athletes are almost forced to do the same or see their standings plummet. It's the same kind of vicious cycle.
The ever-increasing reliance on standardized tests for evaluating student progress and teacher effectiveness has created a cottage industry for "test score enhancement techniques," with each school and each classroom its own little cottage. Whatever positives come from regular evaluations by testing — and there are positives — are negated by the corrupting influence of the tests, both in determining school curriculum and in making teachers and administrators into schemers, cheaters and liars.