This study compared interstate variability of prevalence rates for special education categories from 1984 to 1985 through 2001 to 2002, using the coefficient of variation (CV), which is designed to compare variances when the means of the groups compared are radically different. The category of learning disabilities, presumed by many to be the most variable, was consistently the least variable. Furthermore, as a group, high incidence categories were less variable than low incidence categories. The results strengthen the conclusion of Hallahan, Keller, & Ball (1986) that using state-to-state variability of disability prevalence rates as justification for criticizing learning disabilities identification practices is largely unfounded. Policy implications for methods of learning disabilities identification are discussed.

No abstract