|Back to October Ed Reporter|
|NUMBER 189||THE NEWSPAPER OF EDUCATION RIGHTS||OCTOBER 2001|
|Exclusions Mar 2000 NAEP Math Gains|
WASHINGTON, DC - The 2000 NAEP math test scores show increased improvement among U.S. 4th- and 8th-graders during the past decade. A report released in August includes both national and state-level test results. Scores were slightly higher for 4th- and 8th- grade students in 2000 than they were in 1990, '92, or '96. The news was less encouraging for 12th-graders: Their average score increased between 1990 and 1996, but declined between 1996 and 2000. The average score for 12th-graders in 2000, however, was higher than in 1990. (See graph.)
Experts point out that this news is tempered by the fact that a majority of U.S. students are still not proficient in math. Only 23% of 4th-graders, 22% of 8th-graders and 14% of 12th-graders scored at the "proficient" level. Only 3% of 4th-graders, 5% of 8th-graders and 2% of 12th-graders scored at the "advanced" level. The 2000 NAEP results further show that performance gaps among white students and their African-American and Hispanic classmates have not narrowed.
Some experts believe that "score inflation" due to the exclusion of "learning disabled" (LD) students (i.e., students who are excluded from taking the tests altogether or whose scores are not reported because of "learning disabilities") plays a major role in the state-level 2000 NAEP math results. Revisions to the Individuals with Disabilities Education Act (IDEA) in 1997 required that LD students be given special accommodations during testing at all levels, including the NAEP. These accommodations may include having test questions read to them, allowing extra time for test-taking, or providing "scribes" to write students' answers for them.
When accommodations are not allowed, LD students cannot participate in testing, and because many of the accommodations currently allowed in some states are incompatible with NAEP procedures, there has been a noticeable upward trend in NAEP exclusions of LD students in a number of states.
According to education policy researcher and NAEP expert, Richard Innes, the worst case rate of increase for exclusions on the 2000 NAEP math tests were "noticeably larger than on the 1998 NAEP reading tests." This means the integrity of the reported gains is compromised. For example, North Carolina's increased scores were praised by the state school board and the media. Yet, according to Innes, North Carolina had the largest increase in exclusion ever: 9% more of the state's 8th-graders were excluded in 2000. Nearly one of 10 North Carolina children who could have taken the test in 1996 was excluded in 2000.
Innes maintains that most of North Carolina's 8th-grade score increases between 1996 and 2000 "can be explained solely by the huge number of students excluded from taking the test. The results are largely bogus," he says.
In his own state of Kentucky, Innes estimates that, allowing for exclusions, 4th-grade math scores probably declined about one point, with 8th-graders' scores remaining constant.
Among the states Innes believes may have posted genuine gains are California, Louisiana and Alabama. "California actually reduced 4th-grade LD exclusions by 3% of the raw sample, while posting a 5-point score increase," he explains. "California 8th-graders did not perform well, which might be explained by the fact that the state began a back-to-basics curriculum in the late 1990s. It may have been too late for 8th-graders to overcome their earlier weak instruction."
"Louisiana also mended its ways," Innes adds. "The state had a big increase in exclusions on the 1998 NAEP reading test, but Louisiana's 4th-grade exclusion rate on the NAEP math test was constant between 1996 and 2000. Therefore, Louisiana's 5-point score rise was noteworthy. Louisiana also posted a large score increase among 8th-graders while the exclusion rate hardly varied."
In Alabama, 3% fewer 8th-graders were excluded from the raw sample, but scores averaged five points higher. Innes explains that, for each additional 1% of students excluded from the tests, there is evidence that scores will increase 1-1« points if there are no other changes.