Replication Difficulties of Payne, Hall, Cameron, and Bishara's (2010) Multinomial Processing Tree Model of the Affect Misattribution Procedure

Abstract

Payne, Hall, Cameron, and Bishara (2010, Experiment 1) advanced a multinomial processing tree (MPT) model formally specifying an affect misattribution process underlying responses in the affect misattribution procedure (AMP). The model was empirically supported in two studies, suggesting such an approach may provide deeper insights regarding the psychological mechanisms underlying the AMP and increased accuracy in estimating individuals’ implicit attitudes. Given the potential theoretical importance and applied implications of the approach, we attempted to independently replicate Payne et al.’s Experiment 1 MPT model results in two high-powered studies. Despite considerable effort to closely duplicate all methodological details of the original study, overall we were unsuccessful in replicating the original findings. Our results suggest that the psychological processes underlying the AMP are likely more nuanced whereby different processes may be operating in different individuals. We suggest the use of more advanced modeling approaches – at the level of the individual – to understand the likely heterogeneity that exists in the processes underlying responses in the AMP. We urge researchers to exercise caution in using the AMP for theory-testing or applied purposes until further research clarifies such heterogeneity.

Keywords: Affect Misattribution Procedure, Multinomial Processing Tree model, independent replication, mechanism heterogeneity