The role of mnemonic processes in pure-target and pure-foil recognition memory.

Surprisingly, response patterns in a recognition memory test are very similar regardless of whether the test list contains both targets and foils or just one class of items. To better understand these effects, we evaluate performance over the course of testing. Output interference (OI) is the decrease in performance across test trials due to an increase in noise caused by encoded test items. Critically, OI is predicted on pure lists if the mnemonic evidence for each test item is evaluated. In two experiments, participants received accurate feedback, no feedback, or random feedback that was unrelated to the response on each test trial and pure or standard test lists. When no feedback was provided, performance was nearly identical for standard and pure test lists, replicating previous findings. Only in the presence of accurate feedback were participants able to successfully adapt to pure list environments and improve their accuracy. Critically, OI was observed, demonstrating that participants continued to evaluate mnemonic evidence even in pure list conditions. We discuss the implication of these data for models of memory.