Dr. Gary W. Phillips
Acting Commissioner of the National Center for Education Statistics

September 27, 1999

The purpose of this statement is to bring you up to date on an issue that arose in connection with National Assessment of Educational Progress (NAEP) 1998 reading results for the states, released March 4, 1999. At the time, it was observed that certain states that experienced large gains in reading scores from 1994 to 1998 also had somewhat higher rates of exclusion of students with disabilities relative to other states. The increases in both performance scores and exclusion rates in these states led to debate about the validity of the gains in reading performance in those states.

On May 14, 1999 the previous Commissioner of Education Statistics, Pascal D. Forgione, Jr., issued a statement summarizing research that had been conducted to date, and indicated that, if necessary, NCES would complete a second phase of analyses to examine the possible impact of the change in exclusion rates over time on NAEP reading scores. Additional information we have received, however, has convinced me that additional analyses would not add appreciably to our understanding of the situation.

I would like to share with you the results of recent research that has influenced our decision.

In the previous research, the Educational Testing Service, one of NAEP's contractors, determined that larger rates of exclusion probably did not account for the increase in reading scores in most states.

For example, for one state, Kentucky, ETS found that under differing sets of assumptions regarding the ability levels of excluded students, that state's reading gains could have remained statistically significant or become statistically insignificant if their exclusion rate had remained the same in 1998 as it was in 1994. A separate analysis has now been undertaken by Dr. Lauress Wise of the Human Resources Research Organization (HumRRO) to analyze the possible impact of exclusion on reading performance in Kentucky. The results of those analyses are described in the attached document. Using data unavailable to ETS, Dr. Wise concluded that the impact of exclusion on Kentucky's reading gain, if any, was relatively small. His research provides additional indications that if Kentucky's exclusion rate had been as low in 1998 as it was in 1994, the state would still have experienced statistically significant gains in average reading performance.

NCES is presently sponsoring research into other important questions stemming from our commitment to increase the participation of students with disabilities and limited-English-proficient students in NAEP. These issues include the validity of specific accommodations, how to provide appropriate accommodations to students with varying needs in a large-scale assessment, and whether diverse state inclusion practices have an impact on the validity of NAEP as a nationally consistent yardstick of academic achievement.

Link To:
Part 2 of Remarks