The Eduwonks rightfully tweak the NYT for taking a study with a small, limited sample and running it as research that generalizes to the entire student population:
Ordinarily, when an organization releases a study with the caveat that its sample is "not nationally representative" a national news organization wouldn't then run a big story on it as somehow indicative of a national trend. But not The New York Times when it's education and chance to pop No Child Left Behind in the nose.This new study, while actually very interesting, is not as negative as the NYT story or headline indicate, and is not nationally representative because about 75 percent of the sample is from just four states. In addition, urban districts are underrepresented as are African-American students (substantially). Also, in 7 of the 23 states that make up the sample, only one or two school districts even participate.
To be fair, the NYT article does note the criticisms of the study, and also notes that the results conflict with other existing research. On the other hand, though, the article is critical of NCLB (even thought it's far too early to see when it works, for whom it works, and why), and focuses heavily on the minority achievement gap, even though the study undersampled urban students.
Posted by kswygert at April 14, 2005 12:23 PM