Descriptive and experimental analyses of variables maintaining self-injurious behavior.

Independent descriptive (correlational) and functional (experimental) analyses were conducted to determine the extent to which the two methods would yield data supporting similar conclusions about variables maintaining the self-injurious behavior (SIB) of 6 subjects. For the descriptive analyses, subjects were observed in their residences and at training sites at various times each day while observers recorded naturally occurring sequences of specified subject and staff behaviors. The subjects also participated in a day program for the assessment and treatment of SIB, in which they were exposed to functional analyses that manipulated potential maintaining variables in multielement designs. Both sets of data were analyzed via conditional probabilities to identify relevant antecedent and consequent events for subjects' SIB. Using outcomes of the experimental analysis as the standard for comparison, results indicated that the descriptive analysis was useful in identifying the extent to which SIB was related to social versus nonsocial contingencies, but was limited in its ability to distinguish between positive and negative reinforcement (i.e., attention versus escape).