This exploratory study examined an existing data set containing outcome scores of adolescents and adults on the autism spectrum who were enrolled in 18 different outdoor behavioral health programs in the United States, all of whom are members or cooperating programs of the National Association of Therapeutic Schools and Programs (NATSAP) Practice Research Network. The dataset contained demographic characteristics and self-reported psychosocial functioning of the participants at intake and again at discharge. Data suggest that the self-reported outcome scores for both adolescents and adults demonstrated statistically significant change after the intervention, but when examined for clinical significance, adolescent paired scores had a 49% probability of membership in the Clinical Change Categories of "Unchanged," or "Deteriorated."

Data were further analyzed to determine if clustering patterns between the variables under study could predict membership in one of the Clinical Change Categories. While some clustering effects were noted, results were inconclusive, but highlighted variables on which to focus for further analysis. Multinomial logistic regression results found that a model with the client-reported variables of effort into the program, change in exercise from before the intervention to after, quality of the program, and problem abatement (adolescents only) had significant p-values for predictability of the clinical change category membership of the individual.