Can a five-minute verbal sample predict the response to day psychiatric treatment?

We attempted to predict improvement during day psychiatric treatment by the use of content analysis of 5-min verbal samples obtained at the time of admission of 62 children. As predictors we used verbal sample scales of Hope, Human Relations, and Cognitive Impairment. As outcome measures we used changes on the Severity Level, Aggression, Inhibition, Neurotic, Rare Deviance, and Academic Scales of behavior checklists. We also divided the total group of children into groups classified as organically impaired and not organically impaired by use of neurobehavioral signs. We discovered that verbal sample analysis was only a modestly accurate predictor of improvement, that the Cognitive Impairment Scale was the best predictor, and that it predicted best with not organically impaired children.