A paradoxical dissociation in the effects of midazolam on recollection and automatic processes in the process dissociation procedure

This study used midazolam-induced amnesia to explore the plausibility of the estimates provided by the process dissociation procedure (PDP), which is designed to estimate the contributions of recollection (R) and automatic (A) processes to implicit memory performance. In a double-blind, placebo-controlled, within-subject design with 24 participants, single midazolam doses were administered intravenously, and word stem completion performance was used to calculate PDP estimates. A dissociation was observed such that midazolam decreased R but increased A estimates relative to placebo. Given that a manipulation that induces amnesia would not be expected to facilitate a memory process, these results add to the accumulating body of evidence suggesting that PDF estimates are not always theoretically plausible. Such evidence raises important questions about the use of the PDP.