False memories – remembering an event which did not occur or remembering details of the occurrence differently – has been increasingly researched in the last few decades (Roediger & McDermott, 1995). A well-known method to investigate false memories is through the use of a very simple yet powerful list-learning paradigm which was pioneered by Deese (1959). In this list-learning paradigm, participants are presented with word lists which are all semantically related to a non-presented word, the critical lure (Deese, 1959). For example, for the non-presented critical lure sleep, the following semantically related words were presented: bed, rest, awake, tired, dream, wake, snooze, blanket, doze, slumber, snore, and nap (Deese, 1959). Deese (1959) found that when participants were given a free recall, after the presentation of the word lists, they falsely recalled the non-presented critical lure more often than other non-presented unrelated words. Consequently, he established that this procedure would cause participants to recall words which had never been presented on lists (Deese, 1959). Roediger and McDermott (1995) revived the interest in this paradigm to investigate the effect of false memories experimentally.
Roediger and McDermott (1995) conducted a pivotal study with the primary purpose of replicating and extending on Deese’s (1959) findings. During the first experiment, they gave participants a recognition test after recalling each of the lists which had been presented throughout the study phase. Despite being instructed not to guess, participants confidently recalled and recognised the non-presented critical lure at a higher rate than other non-presented related words (Roediger ; McDermott, 1995). Additionally, they conducted another experiment to gain further understanding on the false-memory effect observed. In particular, they were interested in participant’s phenomenological experience when recognising these non-presented critical lures. To examine this, they applied Tulving’s (1985) procedure, in which participants were asked to make either a ‘remember’ or ‘know’ judgement for each word they recognised. A ‘remember’ judgement was defined as being able to distinctly recollect that the word had been presented, for example, recalling its position on the list (Roediger ; McDermott, 1995; Tulving, 1985). In contrast, a ‘know’ judgment referred to being confident that the word had been previously shown, although participants had no recollection of its actual presentation (Roediger ; McDermott, 1995; Tulving, 1985). Roediger and McDermott’s (1995) findings revealed that participants remembered the non-presented critical lure 72% of the time. These results suggest that false memories for critical lures and accurate memories for studied items are experienced similarly, as participants indicated that they could vividly relive the presentation of the critical lure (Roediger & McDermott, 1995). Presenting participants with semantically associated word lists to investigate false memories has come to be known as the Deese/Roediger–McDermott (DRM) paradigm.
To explain why these robust false memories emerge within the paradigm, many theoretical accounts have been put forward. Two key accounts are fuzzy-trace theory and activation/source monitoring. The former suggests that when people experience an event they simultaneously encode a verbatim and gist representation (Brainerd & Reyna, 2002). Memory traces which correspond to specific information about the studied words, such as its position on the word list, are encoded as verbatim representations (Brainerd & Reyna, 2002; Payne, Elie, Blackwell, & Neuschatz, 1996; Wade, Rowthorn, & Sukumar, 2017). Whereas memory traces about the general semantic content, pattern and meaning of list items are encoded as gist representations (Brainerd & Reyna, 2002; Payne et al., 1996; Wade, et al., 2017). When we study word lists, participants encode both a verbatim representation of each of the items presented and a gist representation, which is general information about the association shared among words on the list (Payne et al., 1996; Wade et al., 2017). As all studied items are semantically related to the non-presented critical lure, which is consistent with the common theme of the list, using gist traces to cue other words is likely to result in false recall/recognition of the critical lure (Payne et al., 1996; Wade et al., 2017). Therefore, according to fuzzy-trace theory, false memories are created because gist traces are used to judge if an item was presented (Brainerd & Reyna, 2002; Payne et al., 1996).
The second account, activation/monitoring framework, posits that when we read each word during the study phase, words related to this concept are activated in semantic memory (Colbert & McBride, 2007; Wade et al., 2017). As the critical lure is semantically related to the words on the list, it becomes activated every time we read a word from the list (Colbert & McBride, 2007; Wade et al., 2017). When participants are later tested in a recall and/or recognition test, they need to monitor and determine the source of this activation (Johnson, Hashtroudi, & Lindsay, 1993). They must decide whether this activation was externally generated by observing the critical lure on the word list, or if it was internally generated by thinking of the semantically related words presented on the list (Colbert & McBride, 2007; Johnson et al., 1993; Wade et al., 2017). This process is called reality source monitoring (Johnson et al., 1993). If the activation of the critical lure is wrongly attributed to having seen the word previously, a failure in reality source monitoring occurs, and consequently a false memory is created (Colbert & McBride, 2007; Johnson et al., 1993; Wade et al., 2017).
Given that psychology is experiencing a replication crisis many studies, including those in the field of false memories, are being replicated to examine the extent to which classic findings exist. Since its emergence, the DRM has received a lot of attention from many researchers who have attempted to manipulate the task conditions to explore the extent to which these robust false memories persist. For example, manipulating presentation format (Toglia, Neuschatz, & Goodwin, 1999), study/test repetitions (McDermott, 1996; Watson, McDermott, & Balota, 2004), study instructions (Gallo, Roberts, & Seamon, 1997; McDermott & Roediger 1998; Neuschatz, Benoit, & Payne, 2003), and delay (Colbert & McBride, 2007; Payne et al., 1996; Thapar & McDermott, 2001). Although research has shown that this paradigm is robust in a variety of manipulations, it would be interesting to assess to what extent this paradigm can be used twice with the same person. Could it be that once you experience this illusion and are aware of how it works, you are less likely to create these memory illusions again?
The present research sought to determine whether performance on the DRM changes when people complete the DRM task a second time. Consequently, participants completed the task on this list-learning paradigm under standard conditions and were later debriefed about the nature of the paradigm, and its ability to induce false memories. They were given a sample list to demonstrate the type of memory illusions they may have experienced. The debrief served as a forewarning instruction for the retest. Moreover, we wanted to determine whether experience and forewarning instructions in the form of a debrief would result in a change in performance when retested after a delay period of either a week or a month. To our knowledge, no one has determined whether providing a forewarning instruction can attenuate the false memory effect after a delay interval.
Research to date has assessed the effects of informing participants about the nature of the DRM task by giving them explicit warnings about the false memory effect (Gallo, 2006; Gallo et al., 1997; McCabe & Smith, 2002; McDermott & Roediger, 1998; Neuschatz et al., 2003). However, in these studies forewarning instructions were given to participants either immediately before or after the presentation of word lists. In those studies in which explicit warning was given about the ease with which participants falsely recognise the non-presented critical lure, before the presentation of word lists, the false memory effect was reduced (Gallo, et al., 1997; McCabe & Smith, 2002; McDermott & Roediger, 1998; Watson et al., 2004). However, it is important to note that this reduction was not sufficiently high to eliminate false memories. It has been suggested that forewarning before study, allows participants to strategically monitor the source of their memories (Gallo, et al., 1997; McCabe & Smith, 2002; McDermott & Roediger, 1998; Neuschatz et al., 2003). By strategically monitoring the source of memories, participants are encouraged to identify the non-presented critical lure as missing during the presentation of word lists, thus reducing the likelihood of false memories (McCabe & Smith, 2002; McDermott & Roediger, 1998; Neuschatz et al., 2003).
In contrast, forewarning participants after the presentation of word lists, but before the memory test, had little or no effect on participants’ illusory memories (Anastasi, Rhodes, ; Burns, 2000; Gallo, Roediger, ; McDermott, 2001; Neuschatz, Payne, Lampinen, ; Toglia, 2001). The high level of false recall/recognition observed in these studies demonstrates that the information that results in false memories, such as the activation of the critical lure or reliance on gist representations, is outside the limits of participants’ conscious control (Gallo et al., 2001). Therefore, if participants are not given any information concerning the characteristics of memories or strategies to facilitate monitoring, it is highly unlikely that they will be able to avoid the creation of false memories (Gallo et al., 2001).
Given that past research has demonstrated that forewarning prior to study causes participants to become more resistant to false memories, the present research sought to determine whether this effect would still be observed when forewarning was provided one week or month prior to the presentation word lists on the retest.
Therefore, the present research examined whether performance in the DRM changes when people complete the DRM task for the second time, after either a week or a month. The extent to which participants are able to use prior experience on the task and warning instructions, will allow participants to effectively monitor the source of their memories, thus becoming more resistant to memory illusions. However, if participants are unable to remember how false memories are induced within the paradigm, forewarning and experience will have little or no effect on the subsequent recognition of critical lures on the retest.
To assess these questions, we conducted an online study in which participants were first instructed to study lists of semantically related words. Proceeding the study phase, participants were given a yes/no recognition test and asked to make confidence judgements regarding the answers they gave on the recognition test, on a scale from 1 (not at all confident in their decision) to 5 (almost certain in their choice). These confidence ratings would allow for a greater understanding of participant’s memories. Once they completed the recognition task, participants were fully informed in the debrief about the nature of the DRM paradigm and its ability to induce memory illusions. They were presented with a demonstration and instructions on how false memories are elicited. Following the same procedure outlined above, one group of participants completed the retest after a week and the second group after a month. Based on findings from past research, we predicted that participants would be able to reduce the false memory effect when completing the retest after either a week or a month, as a result of forewarning and prior experience.