I’m not the only one raising questions about uneven exclusion rates in the National Assessment of Educational Progress for students with learning disabilities.
Education Week’s Catherine Gewertz raises the issue in her March 25, 2010 blog item titled “Exclusion Rates in NAEP: A Pot Still Simmering.” You may have to scroll down a bit to find this as that blog gets more posts.
As Ms. Gewertz writes, the exclusion issue is indeed still simmering, and Kentucky has been at the epicenter of the discussion since 1999. At that time, I wrote a rather detailed paper about the situation. Using simple statistical analysis techniques, I found that the changes in exclusion that occurred in Kentucky and several other states could have resulted in score inflation that ran somewhere between slightly more than half a NAEP Scale Score point for each one percent increase in exclusion to as much as a full NAEP point for each one percent exclusion increase. In Kentucky’s case, where the exclusion rate reportedly rose from 4 percent of the raw sample to 10 percent between 1994 and 1998, there was evidence in the regression that all of the state’s six point score rise might have been due to nothing more than the exclusion rate change.
With NAEP reading score increases running all but flat, those exclusion-based score problems could be important.