DRM in context : story context and the false memory effect in the Deese-Roediger-McDermott paradigm
Michael Lee · Mspace (University of Manitoba) · 2009
demonstrates a way to assess the prediction that a story context that relates to the overall theme of a single DRM list enhances false recognition of the critical, nonpresented word for that list of associated items.The results of Experiment I showed a clear story effect whereby false alarms to critical lures were stronger following the presentation of a DRM story context than after the presentation of a scrambled, non-semantic version of the same story context.Three additional experiments are reported here that further investigated the hypothesis that narrative text Abstract processing enhances the false memory effect in the DRM paradigm.Experiment 2 replicated the first set of results by manipulating context within-participants.Experiment 3 introduced an instructional manipulation so that some participants were given thematic versus list-based vl instructions.In addition to replicating the results of the f,rst two experimênts, the results of Experiment 3 found a clearbenefit across context of thematically-based instructions.In the fourth experiment, a levels-of-text-processing manipulation was provided to participants before the presentation of the DRM stories in their narrative format.Depth of processing had a clear effect on false recognition in that study.Overall, the results of this investigation support a story effect with a story context as a novel experimental variable that appears to moderate the standard DRM effect.The results are discussed in terms of two of the principal theories put forth to explain the DRM memory illusion: the activation/monitoring account and the fiizzy trace theory account, in addition to the potential for using DRM narratives, rather than DRM lists, as a method to explore false memories as produced by more naturalistic materials.vl¡ Ever since Bartlett (1932) first began empirically investigating what he termed reconstructive memory, as opposed to reproductive memory, the scientific study of memory has continued to explore the malleability of memory.Bartlett's work was in many ways a reaction to the traditional approach used by Ebbinghaus, whose early experimental investigations of memory involved materials that invoked simple rote reproductive strategies, such as lists of nonsense syllables.It was Bartlett's belief that memory is a constructive process and that the DRM in Context 2 DRM in Context: Story Context and the False Memory Effect in the D ees e-Roedi ger-McD ermott P aradi gm comprehension, assimilation, and remembering of information conveyed through more natural materials, such as stories or prose passages, was guided by one's organized mental structures orschemata.Since Bartlett, investigations of false memory have focused primarily on thematic material, guided by the assumption that reconstructive memory processes are more likely to be invoked and that false memories will be more likely to occur.Likewise, it was assumed that learning lists of words would lead to reproductive memory processes, and hence, fewer, if any, false memories.One notable exception to the emphasis on reconstructive memory has been the Deese- Roediger-McDermott (hereafter, DRM) paradigm which places the study of false memory within a more traditional list-learning framework.By creating false recognition or recall of words not presented in lists, DRM memory researchers have effectively demonstrated the conditions under which reproductive memory becomes reconstructive, perhaps to the point of making the distinction between the two memory processes unnecessary (Roediger & McDermott, 1995).In a report that has perhaps become better known during the last decade then at the time after its publication, Deese (1959) investigated the probability of extralist intrusions in an