Saturday, December 13, 2014

Chicago Schools Caught Cooking the Charter Books

Back in September, Chicago Principoal Troy A. LaRaviere used the Chicago Public School systems own MAP test numbers to show that public school students were outpacing charter schools in the same neighborhoods.* The findings were published in the Sun Times, complete with linkage to the CPS website where the numbers were all laid out. The Sun Times conducted and published their own analysis, confirming LaRaviere's findings.

Then a funny thing happened. The numbers changed.

The Administrator’s Alliance for Proven Policy and Legislation in Education (AAPPLE) discovered "at some point between the publication of our findings and the release of school ratings, CPS removed the original file containing school growth data and replaced it with a different version." Fortunately, they had saved the original, so they could see the differences that had mysteriously appeared. You can read their report here. But the basic scoop is this;

The main fiddling occurred with pre-test scores, generally lowering them so that school growth would be more awesome. AAPPLE found that this change was made for "nearly every charter school" while fewer than twenty public schools were affected (yes, I hear you out there hollering "but charter schools are public schools" and all I can say is, do shut up). The altered scores gave some charters growth scores increased by as many as fifty points (the biggest change for the public schools was a whopping two points).

Not all charters were winners. CICS chain charters were big winners in up-scoring, while schools like Shabazz Charter took a negative twenty point hit. AAPPLE notes that the charters that were given the fake growth results were largely in gentrifying areas.

AAPPLE did ask for a meeting with the CPS Accountability Office (another exhibit in the How Accountability Makes Education Way More Expensive display) and got one very quickly, in which the office explained that yes, they did change the scores. They offered two justifications.

First, that charters took the 2013 MAP (the "pre-test" or baseline) in the fall of 2013 instead of the spring of 2013. Why this requires a fiddling of scores is unclear, possibly because there is no earthly reason for it.

Second, students took different versions of the same test, so, adjustments are-- really? The AAPPLE report gives the response to these lines of argument, but the bottom lines, as near as I can tell is that only one of two things can be true here. Either 1) the MAP test and growth model system is such a wretchedly invalid system that a stiff wind off the lake is enough to throw its results into question or 2) CPS decided to cheat in order to make some charters look successful. If there's a third possible explanation, I can't see it.

AAPPLE concludes, with what I would call admirable restraint, that lack of transparency is a bit of a problem, and CPS is not using a level playing field. Since I don't have to work in Chicago, I can go ahead and make my own assessment, which is that this is the kind of lying, cheating bullshit that you can only get away with when you work behind closed doors.

* Personally, I am not inclined to evaluate any schools anywhere based on standardized test scores, but that's the game reformsters choose to play, and I do think it's not unfair to judge whether they are winning in the game they chose under the rules they set.


  1. Take a look at the current levels of morality and ethics in big business today in this country. Charter schools and test developers are big businesses. Why would anyone expect them to be on the up-and-up? And is there any wonder why so many politicians are cozying up to them? Politicianss flock to where the money is.

  2. Yes, Virginia, there are still people that choose to teach only because it is the right and moral thing. I steadily grow more despondent though that anything can stop the tide of profit-seeking.

  3. The more I hear, the more my neck hurts from shaking my head in disbelief. However, Bob's name made me laugh, so I feel a tiny bit better for a moment. Thanks, Bob!

  4. If everything I have been taught by the MAP trainers that taught us how to give and analyze the MAP test last year is correct, then neither of the stated reasons should be valid with an adaptive assessment. Of course it was a different test each time children took it-it's designed to be a completely different test EVERY time they take it. The whole point of giving it every year (according to our district and the MAP people), is to be able to compare achievement over YEARS, but just from fall to spring. If the test is changed that dramatically from year to year, then even that marginally beneficial aspect of giving it at all is invalidated. I call BS!