Further evaluation of functional analysis screening methods in early autism intervention.

Abstract

A goal of some functional analysis (FA) variations is to reduce assessment time while still maintaining efficacy. This may be especially important when conducting FAs in early intervention programs, where time is a crucial commodity. To that end, we evaluated a model for using the results of the no-interaction condition as a screening for behavioral function and to guide selection of FA test conditions with 20 participants (22 assessments) aged 3 to 7 years old. We used the no-interaction condition to develop hypotheses for both automatic reinforcement and socially mediated reinforcement. The outcome of the no-interaction condition guided the selection of test conditions for the remainder of the FA. We also incorporated methods from prior FA studies (e.g., divided attention) to modify the test conditions. We obtained differentiated results in 91% of assessments, all within 70 min and, as such, extended evidence that an FA can be completed in little time without sacrificing efficacy.