Tuesday, December 7, 2010

The Urban Prep Myth (or How to Fool the Public by Abusing Data)

My Principal recently sent a memo extolling the virtuous Urban Prep Academy, in Chicago, which has miraculously been able to get low income youth to graduate and attend four-year colleges at rates that would make a statistician guffaw incredulously. The point was, of course, for us to buck up and try harder to make our own students succeed, because if they could do it, then by god, we could, too. The stats, of course, were doctored. Check out the recent post by This Week in Ed:

Media: Fact-Checking Urban Prep's "100 Percent" Grad Rate Claim

image from www.voiceofsandiego.org
At long last and after much prodding, a mainstream journalist has found a way to let readers know that Urban Prep's 100 percent graduation rate is based on the percentage of seniors finishing the year (107 of 107), not the percentage of entering fre


  1. If we are to play their statistics game, we see that better "results" correlate with social class and resources. See this comment in the latest PISA tests report:
    Within countries, schools with better resources tend to do better only to the extent that they also tend to have more socio-economically advantaged students. Some countries show a strong relationship between schools’ resources and their socio-economic and demographic background, which indicates that resources are inequitably distributed according to schools’ socio-economic and demographic profiles.
    New York City Eye.

  2. That is very true. If you compare schools within districts or states, the correlation still holds. In my district, there is a 92% correlation between the schools' API and the percentage of kids on free or reduced lunch.