What is the effect of retention interval on accurate and false recollection in the Deese, Roediger, and McDermott (DRM) procedure? Previous researchhas suggestedthat false recall is more persistent than accurate recall, but the recognition results have been inconsistent. In two parametric studies, we tested recall and recognition for the same DRM lists, over retention intervals that ranged from no delay to a 2-month delay. We found that accurate and false memory were diminished by increases in retention interval, false memory persistence was present for recall and recognition, greater persistence for false memory than for accurate memory was more readily observed for recall than recognition, and the highthreshold (P r ), signal detection (d9), and nonparametric (A9) recognition measures differed in their sensitivity for detecting change. The effect of retention interval on accurate and false memory is consistent with expectations from fuzzy trace theory. In the DRM procedure, truth is not more memorable than fiction.

In two experiments involving recall and recognition, we manipulated encoding strategies,attention, and practice in the Deese, Roediger, and McDermott false memory procedure. During the study of auditory word lists, participants listened to the words, wrote the words, wrote the second letter of the words, or counted backward by threes and wrote numbers in time with the words. The results from both experiments showed that, relative to the full-attention hear word condition, the divided-attention write number condition impaired accurate memory, but not false memory. In contrast, the focused-attention write word and write second letter conditions were comparable to the hear word condition in producing accurate memory, yet they were better at reducing false memory. But even after multiple study-test trials, people still falsely recalled or recognized words that they had never written during study. These results are consistent with predictions generated from fuzzy trace theory and the activation/monitoring framework.

In general, hd-PDT was able to produce similar results as sd-PDT. Subgroup analysis revealed superior results with hd-PDT when baseline vision was 20/50 or better, or when there were three or less polyps on indocyanine green angiography.