I have been reading Diane Ravitch’s Reign of Error in which Ravitch attacks many politicized positions on public education. Her core concern is the privatization of public education and Ravitch disputes multiple complaints against public education based on what she feels are faulty data and flawed research methodologies. A favorite data source appears to be the NAEP (the National Assessment of Educational Progress) which she argues has useful characteristics in that it offers International data over time and it is not a high stakes test. Ravitch argues that many claims made about the failures of U.S. are actually failures in addressing needs created by poverty an important variable which schools cannot easily address. I have become sensitized to references to poverty and the NAEP.
Today I noticed a post by Scott McLeod abstracting comments from a lengthy post by Jonathan Lovell focused on income inequality and student achievement. You can read these posts using the links provided here. The point made is that the U.S. sample in international comparisons oversamples low income students and when adjustments are made statistically “re-weighting” the samples from various countries, the U.S. ranking among nations improves greatly (this is the same argument Ravitch makes).
The source cited for this statistical adjustment is a study by economists Carnoy and Rothstien. I have already identified the gist of their findings. However, in looking through their paper, I located something else I found interesting. The authors contend that the “educational policy advocates” made premature judgment regarding what these results meant (arguing the U.S. scored poorly) and a more careful analysis reveals a much more optimistic picture. They are careful to avoid assigning motives for this rush to judgment.
For some reason, however, although TIMSS released its average national results in December, it scheduled release of the international database for five weeks later. This puzzling strategy ensured that policymakers and commentators would draw quick and perhaps misleading interpretations from the results. This is especially the case because analysis of the international database takes time, and headlines from the initial release are likely to be sealed in conventional wisdom by the time scholars have had the opportunity to complete a careful study.