Process dissociation models in racial bias research: Updating the analytic method and integrating with signal detection approaches

Abstract

Process dissociation models are an increasingly important methodology for studying the influence of race on rapid judgments and decisions. While the process dissociation method has contributed many insights into the processes underlying such judgments, there remain several unresolved analytic issues, including which process model best accounts for observed behavior, how best to account for individual differences in process estimates, and how to connect process dissociation estimates to related research stemming from the signal detection tradition. This paper reviews these issues in depth before presenting a simulation study, which allows for a comprehensive examination of the performance of different analytic approaches. Using the information gleaned from this simulation study, I present an updated approach to modeling process dissociation data that will better allow researchers to address questions of interest. Of primary importance, this approach uses a mixed-model framework, which allows for much better handling of individual differences.