In the July 2007 issue of the journal Urban Education, researchers from North Carolina A&T and Fayetteville State universities examined,

the differences in student achievement scores in reading and mathematics in selected public middle schools of choice and in traditional public middle schools with similar demographics and socioeconomic characteristics in a southeastern school district in North Carolina [Cumberland County] during the 1997-1998,1998-1999,and 1999-2000 school years.

In the study, a “choice” school is defined as an intradistrict school, like a magnet school, alternative school, or open enrollment school. (Charter schools were not included.) The researchers found that choice students had significantly higher test scores in reading and math than similar students in traditional schools.

The editors of Urban Education would never have allowed the article to be published with a conclusion favorable to choice, even intradistrict choice. Indeed, the researchers concluded that choice schools may be “skimming off the cream of the crop,” but they failed to substantiate that tired argument with any actual analysis.