Process dissociation procedure: Core assumptions fail, sometimes

Abstract

The process dissociation procedure (PDP) was introduced as a general-purpose tool for estimating the extent to which performance on various cognitive tasks is mediated by processes that are versus are not under the direct control of our conscious intentions. We investigated two of the core assumptions of the PDP that concern the processes that mediate responding under two different test conditions, called inclusion and exclusion. The invariance of familiarity assumption holds that responding based on the familiarity of items is the same in the inclusion and exclusion conditions, and the invariance of recollection assumption holds that responding based on intentional recollection is the same across these test conditions. The subjects in the first experiment studied low- and high-frequency words before receiving a recognition memory test; the results from the high-frequency words contradicted the invariance of familiarity assumption. Experiment 2 was designed to facilitate overall recall test performance while minimising subjects' ability to make source memory decisions. When test performance was analysed according to the PDP algebra, it showed a very low R component together with a substantial F component in conflict with the invariance of recollection assumption and previous evidence and theorising about the