Reading Wendy Kopp’s new book, I came upon a frequently used factoid: that students in poor areas “who do graduate will read and do math, on average, at the level of eighth graders in high-income areas.” The footnotes cite 2005 NAEP data. Usually I see this equivalency used in terms of race: black twelfth graders performing at the level of white eighth graders.
When I was writing Tested, I tracked down the specific data so that I could use it to illustrate the achievement gap. I spoke with three people who work on NAEP. One, a communications specialist, was comfortable with this use of NAEP data. But two psychometricians there confirmed what I was worried about: that the scales for the two tests were not aligned so that you could compare students across tests. So I found other ways to illuminate the gap—and you probably should too, unless the NAEP team has changed the metrics to make this sort of comparison appropriate.
You are right, and I've tried pointing this out in numberous places.
ReplyDeleteThank you for bringing more attention to this quiet common error.