Relative influences of establishing operations and reinforcement contingencies on self-injurious behavior during functional analyses

In the typical functional analysis in which the antecedent and consequent e vents associated with problem behavior are manipulated, the control conditi on involves elimination of both the relevant establishing operation (EO) an d its associated contingency through a schedule of noncontingent reinforcem ent (usually fixed-time [FT] 30 s). In some functional analyses, however, a ntecedent events are manipulated in the absence of differential consequence s, and a common test condition in such analyses also involves the delivery of reinforcement on an FT 30-s schedule. Thus, the same schedule of reinfor cement (FT 30 s) is not considered to be an EO in the former type of analys is bur is considered to be an EO in the latter. We examined the relative in fluences of EOs and reinforcement contingencies on problem behavior by expo sing 6 individuals who engaged in self-injurious behavior (SIB) to four com binations of functional analysis conditions: EO present/contingency present , EO absent/contingency present, EO present/contingency absent, and EO abse nt/contingency absent. Results indicated that the only condition in which h igh rates of SIE were observed consistently was one in which the EO and the reinforcement contingency were both present. Implications of these results for the design of functional analysis test and control conditions are disc ussed.