Agreement Between Descriptive and Experimental Analyses of Behavior under Naturalistic Test Conditions

We compared the results of a contingency space analysis (CSA) of behavior-consequence recordings to the results of functional analysis (FA) test conditions involving antecedent stimuli and verbal statements that both differed from and mimicked those in the natural environment. Three preschool children with autism spectrum disorder participated. During Phase I, we identified teacher-delivered consequences that were contingent on problem behavior from observational data. During Phase II, each child participated in FA test conditions conducted by their teacher or a same-sex experimenter in an ABA reversal design. Two Experimenter FA conditions were identical to the Teacher FA, with an additional condition involving a different task, demands, and verbal statements. Teacher FAs revealed a clear function for problem behavior for two children, whereas the Experimenter FA was undifferentiated. The reverse pattern was observed for the third child. Results of the CSAs were consistent with the Teacher FAs for two children. Implications for generalizing the outcomes of analog FA test conditions are discussed.