A comparison of fixed versus random order functional analyses

Functional analysis methodology is generally effective in determining functions of behavior; however, occasionally results are undifferentiated due to interaction effects. Iwata et al. (1994) suggested that functional analysis conditions be presented in a multielement design with a fixed cycle order of conditions in order to minimize unwanted interaction effects and capitalize on motivating operations. The purpose of this study was to compare results from functional analyses in which conditions were presented in either a random or fixed order, as suggested by Iwata et al. Functional analyses of aggressive and tantrum behaviors were conducted with six male participants, between the ages two and six. The order in which the two functional analyses were conducted was counterbalanced across participants. The fixed order functional analysis yielded clearer interpretations for one participant.