False recognition with the Deese-Roediger-McDermott-Reid-Solso procedure: a quantitative summary.

Abstract

In the Deese-Roediger-McDermott-Read-Solso (DRMRS) procedure, participants study lists of words associated with central concepts (critical themes) that are not on the lists, then their memory is tested. Based on 224 estimates, the rate of False Recognition of the nonstudied critical themes was .59 (95% confidence interval of .56 to .61), which is smaller than the Hit rate of .75 for correct recognition of studied items (95% confidence interval of .73 to .77) but greater than various rates of False Alarms for other nonstudied items (ranging from .13 to .19). Ratings of subjective confidence were similar on Hits and on False Recognitions but higher than on False Alarms, confirming that false recognition was more like correct recognition than like other errors. The results from judgments of feeling of remembering or knowing, from the effects of intervening activities (particularly recall) between study and test, and from the effects of age suggest that False Recognition occurs because the critical theme is activated along with studied items during list presentation and perhaps also during recall. Invoking fuzzy trace theory, it is argued Hits are based on verbatim traces whereas False Recognition is based on gist traces and a failure of source memory. Proposals are made for research.