The Accuracy of New National Scales for Detecting Emotional Disturbance in Children and Adolescents

Abstract

Through a series of multivariate classification studies, we examined the sensitivity, specificity, and overall accuracy of the Adjustment Scales for Children and Adolescents (ASCA; McDermott, Marston, & Stott, 1993) in detecting emotional disturbance. The ASCA, a teacher rating scale normed on a stratified national sample of 5- through 17-year-old youths, was first applied to a sample of 150 nondisabled subjects and 150 subjects who were socially or emotionally disturbed (SED) matched for age, gender, race, and grade level. Discriminant analyses, including single- and split-sample cross-validations, produced sensitivity, specificity, and overall accuracy ratings of approximately 80%, a level significantly above chance expectancy. The discriminant solution maintained similar accuracy when applied independently to preadolescent, adolescent, male, female, White, and African American children. Accuracy also sustained when the scale was used to differentiate the SED subjects from samples of students who were learning disabled, communication impaired, or gifted/talented, and from a mixture of 1,843 nondisabled subjects and 537 others with learning, mental, and physical disabilities. The scale's performance was evaluated in light of classification accuracy research on other popular instruments.