CEPR - Center for Economic and Policy Research


En Español

Em Português

Other Languages

Home Publications Blogs Beat the Press Waiting for Superman to Stop Cheating

Waiting for Superman to Stop Cheating

Monday, 28 March 2011 14:55
USA Today ran a carefully researched article that strongly suggests that much of the rise in school test scores under school chancellor Michelle Rhee was due to teachers' cheating. Teachers had a substantial incentive to cheat since they would get an $8,000 bonus if their students improved beyond set levels. This is the sort of serious investigative journalism that is rarely seen anymore.
Comments (14)Add Comment
written by izzatzo, March 28, 2011 3:27
Test scores are the only meaningful measure of performance available for markets to distinguish between newly graduated winners and losers.

What's next Mr Nanny, removing the points scored in sport contests so cheaters won't have incentives to take steroids?
the meaning of test scores
written by Fempus Tugit, March 28, 2011 4:00
Those who mistake quantities for qualities should not cast stones.

Really, is izzatzo such a Pollyanna that s/he doesn't know that tests primarily test for which socioeconomic class you fall in (here in the states)?
If you cannot make the students smarter
written by Floccina, March 28, 2011 4:02
If you cannot make the students smarter and you cannot make the students learn more which look like facts. Maybe you can get the students to lean more valuable and practical things that will have a bigger impact on helping them live a better life.

Do those test measure anything important. I think not.
written by JTM, March 28, 2011 4:21
izzatzo: There ought to be a corollary to Godwin's Law, for sports analogies. Obviously this is lot more complicated than balls through the basket. Tests can be the "only meaningful measure" if they are in fact meaningful. Tests are measurable. Sometimes that is mistaken for meaningful, but that is a mistake.
Shame on them!
written by Dom, March 28, 2011 4:40
This is just like a police detective frames an innocent man to solve a case in order to get a pay raise. Shame of those teachers! All they care about is themselves. Must be something in the food they eat. The license of those teachers who cheated should be taken away for good.
written by PeonInChief, March 28, 2011 4:49
It's not kind to be catty, but it's been reported that the amazing test scores that Michelle Rhee reported during her short stint as a classroom teacher, uh, could not be verified. In fact, some have argued that the results she claimed were a statistical impossibility.
written by izzatzo, March 28, 2011 6:21
Oh yeah? Godwin's law? This is the drunk Michelle Rhee under the streetlight looking for the car keys because the light's better isn't it.

If you don't know how to measure performance then perform measurements that result in the desired performance.


Stupid vegetarians.
performance measures
written by Fempus Tugit, March 28, 2011 10:15
Performing measurements tend to change what is being measured (even in a hard science like physics). On the other hand, a test made to measure one quantity may by its very construction preclude simultaneous measure of another quantity.

But this still doesn't solve the problem that quantitative measures will never be able to measure quality. When Quality comes along, then sure it will blow the socks off the measuring device. But to mistake the measuring device as the motivation for the quality performance is just faulty logic (or, as izzatzo calls it, stupid). Did Jesse Owens run for the gold medal? No, he ran the fastest because (1) he could and (2) he loved to run faster than the others. So yeah, competition is a motivator, but you have to have talent too, for those gold medals. But we're just talking about being able to walk to the finish line here, not winning the ultimate footrace. Not that Mr. Meat-And-Potato Head could understand that ...
written by Doc at the Radar Station, March 29, 2011 5:50
What about some of the teachers who were likely fired because they didn't cheat? Ethics is just as important as good spelling, isn't it? NCLB should be repealed. We should be wary about the incentives that such laws create.
Retired college professor
written by Crowbar3, March 29, 2011 10:42
I found no mention of erasures from right answers to wrong answers. Did anyone tally these. If ALL of the erasures were from wrong answers to right ones, one would have to conclude that cheating occurred.
written by JTM, March 29, 2011 5:16
One further thought. The objections to the test-centric regime of Rhee, et al. is that it's trying to evaluate teachers without accounting for varying socio-economic environments which affect the receptivity of students.

Rhee seems to be saying that poverty is just an excuse for poor teaching. It's the "soft bigotry of low expectations" per that distinguish educator G.W. Bush. But both common sense and I believe significant research says that that economic status is a (or the) major driver of student performance.

So yes it's valuable to test student achievement. I don't think anyone says otherwises. But to make student performance a direct measure of teacher performance, without accounting for varying student receptivity, is unfair to the teachers -- and it's guaranteed to yield distorted results.

If Hitler made Jesse Owens run his race on wet grass while others ran on a cinder track, you wouldn't think that was a fair measure of performance (if you know anything about running). Performance depends on relevant conditions as well as the performer's talent and effort. That's what "level playing-field" is all about.

Same with teaching. And I don't think anyone knows how to appropriately adjust student results for varying economics in order to accurately measure the teacher's "value added". That's the problem which this whole regime faces right now.
written by Delphi_ote, March 30, 2011 12:08
Crowbar3, they mention the net improvement due to erasures. It's not very clear, but I believe the "12 correct" to "one correct" statistic means the difference in score due to erasures. Since these are multiple choice questions, it should follow a binomial distribution. That p-value extrapolation about winning the lottery seems like it's from that kind of analysis.
written by purple, March 30, 2011 5:46
It's pretty easy to cheat on these tests, or lend some guidance and no one would know.
written by purple, March 30, 2011 5:53
Seriously, there is one adult alone in a room full of busy kids, the kids pass in the test when it is done,etc.

A lot naive people out there.

Write comment

(Only one link allowed per comment)

This content has been locked. You can no longer post any comments.


Support this blog, donate
Combined Federal Campaign #79613

About Beat the Press

Dean Baker is co-director of the Center for Economic and Policy Research in Washington, D.C. He is the author of several books, his latest being The End of Loser Liberalism: Making Markets Progressive. Read more about Dean.