The College Board released the 2012 scores for college-bound seniors last week. Didn’t notice, did you? That’s because there was once again no news: Both the national scores and the scores for different ethnicities didn’t budge by more than a point or two.
It’s been pretty much like that for twenty years now. Following a plunge in SAT scores from the mid 1960s to 1980 and a mild rebound during the 1980s, SAT scores have shown no significant change since 1992 (when the SAT was recentered) for whites, blacks, or Latinos. Only Asian scores have risen meaningfully, from 487 to 518 in verbal (“critical reasoning,” as it is now called) and from 551 to 595 in math, which amount to increases of about .28 and .39 standard deviations respectively. White increases in verbal and math amounted to .07 and .19 standard deviations respectively, while both black and Latino scores were flat on verbal and increased .08 standard deviations in math. (A technical note: My Latino scores combine the SAT scores for Mexican-Americans and Puerto Ricans, reported separately by the College Board, weighted by the proportion of test-takers in each group as of 2011.)
For all four groups in both tests, a linear trendline is the most parsimonious one (although a marginal case could be made for rising white trendlines that have flattened since the early 2000s). For blacks and Latinos, the trendlines are effectively flat.
Can this lack of progress in closing the traditional test score gaps be attributed to a widening pool of test-takers among disadvantaged minorities? It’s the explanation that the College Board has trotted out ever since the gaps were first disclosed in the early 1970s, but what we’re seeing in the SAT is not much different from what we see in the National Assessment of Educational Progress (NAEP), which tests all 17-year-olds, not just those who take the SAT.
The graphs below are based on the National Center for Education Statistics’ Long-Term Trend Assessment. They show the same start year as the graphs for the SAT, 1992, but end with the most recent assessment in 2008. No collateral data suggest that anything dramatic has happened since then.
For the NAEP, the trendlines are even flatter than they were for the SAT, with the largest being a .14 standard deviation gain in reading by Asian/Pacific students. That is also the single set of scores for which a nonlinear trendline may be more appropriate than a linear one.
The point of presenting these graphs is to keep our eye on the big picture. I have no argument with those who explore the different and somewhat more optimistic results from the NAEP tests of younger students, nor with those who seek to document impressive gains by special educational approaches such as the Knowledge is Power Program. But all of those endeavors have to keep in mind the elephant in the corner: Whether we’re talking about college students or the entire population of 17-year-olds, the product coming out of the end of the K-12 system has not improved for the disadvantaged groups that have been the object of wrenching upheavals that Congress has imposed on the public school system over the last twenty years, with No Child Left Behind being the most wrenching and, in my experience as the parent of children in public schools, destructive. The only group that has improved significantly, college-bound Asian students, is the one that has received the least attention by the panoply of special programs to help minorities.
The politicians who will insist that we continue to pay for, and probably expand, the same-old same-old federal K-12 programs in the next session of Congress should have to explain why the famous definition of insanity, doing the same thing over and over in hopes of a different result, doesn’t apply to them.