Behavior states are real and observable.

Mudford, Hogg, and Roberts (1999) published additional concerns about their failure to achieve interobserver agreement on adults with mental retardation when using an experimental, 13-category behavior state code we earlier provided to them at their request. They presented findings of interobserver agreements from a videotape of their 3 adult participants that was scored by observers at the University of Kansas. In this response we present extensive investigations demonstrating the reliable measurement of behavior state among infants and adults with mental retardation, including independent data from other state researchers. We give arguments showing why their "videotape study" does not meet requirements for acceptable research methodology and discuss other procedural issues for measuring state behavior among infants with and those without disabilities.