Below is a comment from Tom Loveless, senior fellow at the Brookings Institution, about my post from yesterday showing big declines for top U.S. students on the PISA math test from 2003-2006. [quote]
Because it does not assess mathematics taught in the school curriculum, the PISA exam is not a reliable source of evidence on the issue of whether high-achieving students have lost ground or made gains during the NCLB era. But it is worth asking whether the trends you highlight from the PISA data are aligned with other respected assessments. On the main NAEP, long-term trend NAEP, and TIMSS, the evidence is mixed.But on a related question--is greater progress being made at the 10th percentile than the 90th percentile?--the evidence strongly suggests that it is. The gap between the highest and lowest achievers is contracting, and that began around 1998 or 1999 depending on the test. That's clear from both NAEP tests. When gains at the top were made, they were often accompanied by even larger gains at the bottom.
This phenomenon did not occur in the early to mid-1990's or in most of the previous history of the long-term trend NAEP (the only one of these tests available before 1990). However, as I explained in Fordham's High Achieving Students in the Era of NCLB report, it could be spotted in the late 1990s in states that already had accountability systems in place. Moreover, it is not as pronounced among private school students as among public school students on the main NAEP. I wrote about this in the recent Brown Center Report.
The evidence suggests accountability systems that reward gains at the bottom of the distribution and ignore the progress of students at the top--systems which pre-date but also include NCLB--are driving the phenomenon.
-Mike Petrilli
