Tuesday, August 31, 2004
Lies, D*** Lies, and Statistics: Charter Schools, AFT, and NAEP
The American Federation of Teachers recently released its analyis of the US Department of Education's National Assessment of Educational Progress. Countless headlines have trumpeted the finding that students in charter schools significantly underperform students in ordinary public schools, which isn't what the AFT report says at all.
The data in the NAEP itself are rather raw and require some real effort to analyze; see for yourself. So I guess I could understand if reporters got the story wrong because they had trouble analyzing the data itself. But that wasn't the problem; they inaccurately reported the contents of AFT's own very understandable report on the data.
It is well known that the AFT is one of the largest, most powerful, and most liberal lobbying groups in the nation, and that they and the educational establishment generally are bitterly opposed to any measure that would contribute to any affordable, serious competition for the public schools. So it would not be surprising if the AFT's study, or its report on its study, were skewed against charter schools. However, you only have to dig as deep as the bottom of page ten in the AFT study to read this:
"When comparing student achievement by race, no meaningful difference existed between charter schools and other public schools. In fourth-grade math, white, black, and Hispanic students in charter schools had lower average scale scores than their peers in other public schools, but the differences were small and statistically insignificant. In reading, the gaps were even narrower."
The following page gives the supporting data.
Here's what that is saying: Charter schools enroll a disproportionately high percentage of inner-city minority students, mostly African-Americans, whose achievement, regrettably, is well known to be much lower than average, for an assortment of cultural, economic, and political reasons. When the charter-vs.-public numbers are adjusted for this, there is "no meaningful difference" between charter and public school student achievement.
So if the study itself didn't say what the media reported, why didn't all those headlines read, "Charter Schools as Good as Public"? There are two major points between the study and the newspaper articles: AFT's press release and the reporters who wrote the stories. It turns out they both went wrong.
We have the press release itself, a sort of smoking gun. It mostly complains that the Department of Education repeatedly delayed releasing the data about charter schools. But it does include the claim, "charter school students mostly underperform and sometimes score about as well as regular public school students" -- a half-truth at best.
In the case of the media, we don't need a smoking gun. Either the people writing the stories didn't bother to read the AFT report all the way to page ten, or they didn't understand it, or they did read and understand it, but didn't want to report something so different from the AFT party line. I'm not sure which of the possibilities is worse: lazy reporters, intellectually challenged reporters, or deliberately deceptive reporters with an agenda.
See commentator Clarence Page on the same point.
Copyright 2004 by David Rodeback.