Metamemory and the word frequency mirror effect

The word frequency mirror effect is the observation that high frequency words produce more false alarms and fewer hits than low frequency words in recognition memory tests (Glanzer & Adams, 1985). Both single and dual process models of recognition memory have attempted to explain this finding. Single process accounts have argued that both hits and false alarms are the result of familiarity (Glanzer and Adams, 1990), whereas dual process accounts have argued that hits and false alarms are dissociated from each other, with hits being the result of both familiarity and recollection (Joordens and Hockley, 2000). In six experiments, the impact of metamemory judgments on the mirror effect was investigated. Results demonstrated that metamemory judgments affected the hit portion of the effect while the false alarm portion of the effect remained consistent across experiments. The dissociation provides support for dual process accounts of recognition memory.