Children with Autism Spectrum Disorders (ASD) generally have pervasive social skill deficits, and their social behavior provides a pivotal link to cognitive and communicative outcomes. Researchers in applied behavior analysis have used functional analysis (FA) to identify functions of aberrant behavior and successfully linked intervention to function for children with ASD, but they have not extended FA techniques to determine functions of prosocial behavior. The proposed study thus has 3 aims: To determine (a) whether primary functions of prosocial behaviors in children with ASD can be identified using FA, (b) the extent to which primary functions systematically differ across ASD social subtypes, and (c) the effectiveness of social skill treatments linked to FA outcomes. Using a FA protocol adapted from our preliminary work, we will obtain rates of prosocial behavior for 60 participants ages 3-9 with diagnoses of ASD, 20 for each subtype of autism (active-but-odd, passive, and aloof), in each of 3 FA conditions: obtaining attention; obtaining a preferred tangible item; escaping social stimuli. Each participant will receive the conditions in multiple 5-minute videotaped sessions (one condition per session) presented in random order, and the condition with the consistently highest rate of prosocial behavior will indicate the primary function for a given participant. To determine whether function and subtype are dependent, we will use a 3 x 3 chi squared test; we will use a muiticategory logistic regression analysis to determine whether chi squared findings are maintained after controlling for IQ and communication score. To pilot the effectiveness of treatment linked to function, we will use a within-subjects reversal design with 3 participants with different hypothesized function x subtype combinations. Each will receive all 3 contingent treatments (attention, tangible, escape), and we will thus compare (a) the effectiveness of contingent treatment linked to FA outcome with non-contingent treatment and (b) rate of behavior across contingencies to determine whether function-based treatment (as determined in Aim 1) is more powerful than contingent treatment not linked to function. The outcomes of this exploratory study will guide future research efforts to determine whether function- based social skills treatments can be effectively matched to characteristics of children with ASD. ? ? ?
