Two key effects of the Implicit Relational Assessment Procedure are moderated by procedural and performance variables: potential theoretical implications

Forty-two participants completed an Implicit Relational Assessment Procedure (IRAP), meeting the practice criteria (median latency < 2000ms and accuracy > 80% correct), and thus proceeded to six test blocks. The IRAP had four trial-types: TT1: happy-face/happy-word; TT2: happy-face/fear-word; TT3: fear-face/happy-word; TT4: fear-face/fear-word. In consistent blocks, participants had to respond True, False, False, True, respectively, to the four trial types; responding in the opposite pattern was required in inconsistent blocks. Participants were randomly assigned to start with the consistent (n = 21) or the inconsistent (n = 21) block. Difference (D) scores were calculated (inconsistent minus consistent latencies). A significantly larger D score was observed in TT2 relative to TT3, but only for participants who started with the inconsistent block. Results also indicated a “happiness superiority” effect (i.e., largest D score on TT1), but it was moderated by adherence to the practice criteria; only participants who failed to maintain the criteria across any of the trial types showed a significant superiority effect. Overall, the findings highlight the importance of distinguishing between the orienting/evoking functions versus relational properties of the stimuli presented within an IRAP.