|Back to July Ed Reporter|
|NUMBER 162||THE NEWSPAPER OF EDUCATION RIGHTS||JULY 1999|
NEW YORK, NY - Two states that claimed significant improvement in 4th grade reading scores on the 1998 National Assessment of Education Progress (NAEP) tests would have scored much lower had more special education students been tested as had traditionally been the practice. Education Week (May 19) noted that, had Maryland and Kentucky not excluded large numbers of special ed students during the time period 1994 - 1998, their NAEP test gains "would not have been statistically significant." A third state that posted modest gains may also have been helped by excluding lower-performing students.
The Educational Testing Service (ETS) reviewed the NAEP scores and released its analysis on May 14 in a document entitled "A Summary of Initial Analysis of 1998 State Reading Results." The Associated Press reported on the same day that the chief of the Education Department's statistical branch, Pascal D. Forgione Jr., said that neither the states' test scores nor their rankings would be affected by the discovery of the exclusions, but that "the federal study of the 1998 test results could lead test-givers to change some policies." A few days after he announced the ETS study findings, Forgione resigned his post as statistics commissioner, despite the fact that he was both "popular and respected" (Education Week, May 26).
In February, Forgione had "sided with members of the NAEP governing board who said Vice President Al Gore had jeopardized the credibility of the NAEP by announcing the 1998 reading-test results at a campaign-style event." Board policy calls for the scores from the federal testing program to be released by the statistics commissioner (Education Week, May 26). Interestingly, the White House had refused to support Forgione's renomination to the post based on his "failure to meet income-tax deadlines for eight consecutive years," though there were no actual violations of federal tax law.
Of the more than 30 states participating in the NAEP tests, at least 18 "left out more special education students from the 1998 testing sample than from the 1994 group." Kentucky excluded 10% of these students in 1998, compared to 4% in 1994. Connecticut removed 10%, compared to 6% in 1994, and Louisiana removed 13%, compared with 6% in 1994 (Associated Press, May 14).
On May 17, USA Today reported that the Department of Education is planning "more investigations into the accuracy of recent national reading test scores." The article noted that the number of excluded students "becomes key as more [students] are being put in special-education programs. From 1994 to 1998, the special-education populations in Louisiana and South Carolina increased 8.5% and 21%, respectively. A Kentucky education activist quoted in USA Today commented: "We are writing children off instead of teaching them how to read."
A total of 10 states posted large gains in average NAEP reading test scores from 1994 - 1998. After adjustments for the excluded students, eight of those states still showed statistically significant increases. Only Kentucky's gains would have disappeared completely, which is of interest to many in the education field because Kentucky is the home of the educational testing reform system known as KIRIS (Kentucky Instructional Results Information System), which was touted as a first-of-its-kind reform model for other states in 1990. The KIRIS test proved to be so flawed that many parents, teachers, and some legislators demanded it be scrapped altogether. It was revamped last year instead, and is now known as the Commonwealth Accountability Testing System, or CATS (See Education Reporter, June 1998).
As Chester Finn, former chairman of the National Assessment Governing Board (NAGB), summed up the NAEP testing scandal: "It's a serious concern if we're trying to report progress and states are fiddling with numbers of special-education students that take the test. You cannot make fair comparisons [in that manner] over time."