Starting this year, results from the National Assessment of Educational Progress will be based on samples that include special-needs students who take the tests with accommodations.
But a study from the National Center for Education Statistics suggests that policy could pose challenges for NAEP.
The study examines 1998 NAEP reading scores when calculated to include results from special-needs students tested with accommodations, such as more test-taking time. While scores for the nation remained essentially unchanged, it found the same was not true at the state level.
For grade 4, average scale scores were lower in nine states when the sample included students who had taken the test with accommodations. For grade 8, though, including students who were given accommodations made no difference in either national or state results.
"Clearly, the increased inclusion that results from providing accommodations appears to bring into the sample students who are generally low performers," the report concludes. Still, it argues that doing so provides a more representative picture.
In 12 jurisdictions, the percent of students excluded from NAEP rose from 1994 to 1998, based on samples where accommodations were not permitted.
"An important question for the NAEP program is whether the offering of accommodations will offset what appears to be a tendency in some jurisdictions toward increased exclusion," the report says.
Richard G. Innes, an education activist in Kentucky, asserts that improvement in participation rates was less than 2 percent for most states, even after accommodations were offered. He has expressed concern that states with high exclusion rates benefit greatly, and unfairly, in NAEP comparisons.
Sharif M. Shakrani, the deputy executive director of the board that oversees NAEP, said it is imperative to monitor changes in the participation rate in interpreting NAEP results.
—Lynn Olson [email protected]
Vol. 22, Issue 26, Page 10Published in Print: March 12, 2003, as Testing