Performance at the Top

Abstract

Anumber of recent studies have indicated that many districts focus on the bubble or the tippers--kids who with some extra instruction might tip the school into achieving Adequate Yearly Progress (AYP). Jennifer Booer-Jennings found that teachers were told to ignore hopeless cases and sure things (see the December 2005 Research column, Blowing Bubbles in Texas) in favor of improving the scores of kids on the bubble. Does the focus on making AYP and lowering the achievement gap mean gifted kids are losing out? Logically, they can't make the same progress as low-achieving kids or else the gap will never close. Three recent studies throw some light on the issue. Loveless Study of NAEP Scores Tom Loveless of the Brookings Institution examined National Assessment of Educational Progress (NAEP) data in reading and mathematics at grades 4 and 8 for the 10th and 90th percentiles. This column focuses only on fourth-grade data. Among fourth graders, Loveless found that between 2000 and 2007 the 10th percentile scores in math rose 13 points, about a year of achievement growth, while scores for the 90th percentile kids rose only 5 points. Why look at test scores in 2000 when NCLB didn't take effect until 2002? Loveless says that by 2000, school people knew something like NCLB was coming and, practically, there was no later assessment until 2003, over a year after NCLB became law. But recent stories indicate states made plans to achieve 100% proficiency that required little effort in the early NCLB implementation. This is important because for the 10th percentile students, 7 points of the 13-point gain came in that interval between 2000 and 2003. It's impossible to say when the gains occurred, but NCLB seems unlikely to be responsible for them. Fourth-grade reading scores are even more problematic. The 90th percentile score went up only 3 points over a seven-year period. The 10th percentile rose 16 points, but most of that-12 points--occurred between 2000 and 2002-again, before NCLB was actually in place. Loveless says that he uses a conversion: 11 NAEP scale points = one year of academic growth. Some researchers use 10, some 12, but if we take Loveless' 11, the difference in reading between the 10th percentile and the 90th percentiles in 2007 was 90 points, or 7+ years of growth. This is pretty close to the results in 1990 when the difference was 82 points. These figures become more ominous when we look at the data between 1990 and 2000. For math, the 90th percentile rose 12 points and the 10th percentile 13. In reading, the 90th percentile rose by only one point, but the 10th percentile lost 11 points. The 10th percentile bottomed over this time in 1994, and the score is up 1 point from 2000. So, for reading at least, NCLB has not acted as Robin Hood, taking from the rich and giving to the poor. However, the data overall do seem to corroborate an unpublished report from Marshall Smith of the Hewlett Foundation that NAEP scores have grown more slowly after NCLB became law than before it. Loveless uses the average per year gain and thinks it's faster after NCLB. But as I pointed out, most of the gains came before NCLB was really in place. I argue that the trend line paints a better picture than average gain or loss per year. States that were deemed to have accountability programs showed more gains in math for their 10th percentile than for their 90th percentile. Non-accountability states showed little difference in gains. In both accountability and non-accountability states, the 10th percentile rose more than the 90th. The 90th percentile showed little change in either. The 10th percentile included more black students (36.9%) and about equal numbers of white (28.4%) and Hispanic students (29.8%). The 90th percentile students were largely white. Only 16.1% of the teachers teaching 90th percentile kids had 0-4 years of experience, compared to 29. …