I once sat on a committee whose task was investigating cheating on statewide assessments. Also on the committee: State Department of Ed assessment specialists, legislators, business leaders, school administrators and a couple other teachers. Assigned seating at round tables was designed to ensure mixed perspectives in discussions. After a presentation which--pretty convincingly-- indicated that some schools (identified only by number) were cooking the answer sheets, we were supposed to provide recommendations for what to do about it.
Then the fun began.
The teachers and administrators immediately focused on why the cheating occurred: what would push teachers or principals to take the huge professional risk of erasing answer sheets, simultaneously invalidating useful information? There was an immediate assumption that schools where cheating was obvious were trying to shield children from bad news about their life prospects, and let hard-working educators save face--a misguided and unethical, but explicable, response to high-stakes testing.
The legislators and business leaders, on the other hand, went directly to consequences. Their assumption was that adults were cravenly conspiring to hide their incompetence to keep their jobs. They were mostly interested in ferreting out the more subtle and creative cheaters who hadn't been picked up by DOE staff.
One of the folks at my table, a high-ranking executive at one of the Big Three, whose name was recognizable from media coverage of stories about the faltering auto industry, proposed an anonymous call-in system. Teachers, parents and administrators would be encouraged, via TV and radio spots, to call the State DOE and share their suspicions about devious colleagues--or their child's teacher. We could call it the 1-800-CHEATER hot line! (No, the number doesn't work. I couldn't figure out how to remove the automatic HTML coding.)
The implications of Michigan children listening to repeated TV ads urging trusted adults to--not to put too fine a point on it--rat each other out for reprehensible behavior wasn't a concern, evidently. And this was long before NCLB--or the resurgence of test-based merit pay discussions stimulated by RTTT and the Blueprint for re-authorizing ESEA.
Interesting story in NY Times ruminating about whether cheating may be on the rise--and why. Alexander Russo thinks the story is unsubstantiated by hard data. Maybe so, but the prima facie evidence makes sense, to someone who has administered statewide tests. Freakonomics suggests the problem is global. Besides--the real issue is not tracking down the precise percentage of schools that cheat. It's determining why administrators and teachers feel compelled to cheat--and figuring out what to do about the impact of high-stakes testing on both student learning and instruction.
At our last committee meeting, a superintendent in the group shared this story: His district had been singled out as a likely cheater. Not because of erasures or other data patterns--but because their scores had leapt upwards more than 50 points in a single year. Over half the students in his district lived in homes where English wasn't the primary language--families that used the Arabic alphabet, in fact.
State law allowed the district to pull out data from ESL students in reporting, but the superintendent had been adamant that data reports should include all students. He wondered how they would let the community know what they were dealing with if they hid their very real challenges--how they could compare satisfactory achievement in native English speakers (a percentile in the mid-80s) with improving instruction and outcomes for ESL students? He eventually succumbed to pressure to report only non-ESL data, and said that the media never questioned their meteoric rise in achievement. But--he was quickly accused of cheating by the State DOE.
Campbell's law strikes again.
Join the discussion on testing at Teachers' Letters to Obama.