کوچک شماری یادآوری قبل و استعداد ابتلا به حافظه کاذب: دو طرف یک سکه؟

In two studies, we explored whether susceptibility to false memories and the underestimation of prior memories (i.e., forgot-it-all-along effect) tap overlapping memory phenomena. Study 1 investigated this issue by administering the Deese/Roediger–McDermott task (DRM) and the forgot-it-all-along (FIA) task to an undergraduate sample (N = 110). It was furthermore explored how performances on these tasks correlate with clinically relevant traits such as fantasy proneness, dissociative experiences, and cognitive efficiency. Results show that FIA and DRM performances are relatively independent from each other, suggesting that these measures empirically apparently refer to separate dimensions. However, they do not seem to define different profiles in terms of dissociation, fantasy proneness, and cognitive efficiency. Study 2 replicated the finding of relative independence between false memory propensity (as measured with the DRM task) and the underestimation of prior memories (as measured with an autobiographical memory dating task) in people with a history of childhood sexual abuse (N = 35).

Significant life experiences tend to be well remembered. Some people, however, report that they have had the experience of forgetting and subsequently recovering memories of aversive life events. Since the 1980s, many studies have focused on mechanisms that may underlie such recovery of autobiographical memories. Recovered memories refer to autobiographical recollections that are perceived as having been unavailable or inaccessible for some period of time (Sivers, Schooler, & Freyd, 2002). According to Schooler (1999, p. 205), “An individual who reports recovering a memory for trauma is really indicating two sentiments: (a) that abuse occurred and (b) that there was a period of time in which the memory was not available.” Schooler and colleagues were among the first to publish systematic case studies of individuals reporting to have experienced the discovery of long-forgotten memories of abuse (Schooler, 2001, Schooler et al., 1997 and Shobe and Schooler, 2001). They observed that some people reporting recovered trauma memories had mentioned these experiences to their partner, family or friends before the “recovery” of the abuse episodes (Schooler et al., 1997). It appeared that these individuals forgot having remembered the traumatic event, producing an “illusion of amnesia”. Schooler (1999) hypothesized that this underestimation of prior remembering could explain how people may come to believe that they previously forgot an important episode, a phenomenon that he termed the “forgot-it-all-along” (FIA) effect. With this in mind, Merckelbach et al. (2006) examined the existence of this effect for autobiographical details in different samples (see also a recent study by Abenavoli & Henkel, 2009 for a similar procedure). Merckelbach and colleagues instructed undergraduate participants and women with childhood sexual abuse (CSA) experiences to recall vivid memories of certain childhood events. After a delay of 1 h or 2 days, participants were asked whether they had recently thought about any of these events and several others. Despite recalling the events either an hour or 2 days before, many participants claimed not to have thought about these events for years. Relative to individuals reporting continuous CSA memories, those who reported having recovered CSA memories were more likely to forget remembering the recent events during the laboratory tasks. This finding provides indirect support for Schooler’s (1999) idea that the forgot-it-all-along bias is typical for at least some individuals reporting recovered memories of CSA. Another line of research examining memory performance of people reporting recovered CSA memories demonstrates that this group exhibits a heightened vulnerability to develop false memories. A case in point is a study of Clancy, Schacter, McNally, and Pitman (2000) that relied on the Deese–Roediger–McDermott (DRM) task (Roediger & McDermott, 1995). Basically, this task has been shown to be effective in eliciting laboratory induced false memories for semantically related words by requiring participants to study a list of words, all of which are strong associates of a non-presented critical lure. For example, participants are instructed to study words such as injection, sharp, pain, thread, words that are all associated with the non-presented critical item needle. Typically, many participants falsely recall and recognize the non-presented lures on later tests ( Roediger, Watson, McDermott, & Gallo, 2001). Clancy et al. (2000) found that individuals reporting recovered CSA memories more often falsely recognize the non-presented critical lures relative to individuals with continuous CSA memories and control participants with no history of sexual abuse. The pronounced false memory effects on the DRM and the enhanced forgetting of prior remembering in people reporting recovered memories seem to indicate different accounts of recovered memory reports. Clinical vignettes are also suggestive of differing recovered memory experiences (Schooler et al., 1997; see also McNally & Geraerts, 2009). Thus, on the one hand, there are persons who gradually came to believe that they are abuse survivors. They typically attribute current problems to their repressed memories of abuse. The abuse events were mostly recovered step by step, often promoted by suggestive therapeutic techniques such as hypnosis, dream interpretation or guided imagery. It is tempting to assume that these people exhibit a heightened susceptibility to false memories. Another pathway seems evident in people who are suddenly reminded of events and mistakenly believe they have not thought about the events for many years. They experience their recollection as shocking and often describe it as an “Aha-Erlebnis”. Here, it is plausible that individuals have a tendency to overlook prior remembering, i.e., are susceptible to the forgot-it-all-along (FIA) effect rather than to false memories (see for example McNally & Geraerts, 2009). Geraerts et al. (2009) suggest that the FIA account offers a mechanism to explain recovered memory experiences involving events that actually did occur, whereas the DRM account offers a mechanism to explain false recovered memories. However, whether susceptibility to false memories and the FIA effect reflect two different and unrelated pathways has not yet been firmly established. Furthermore, the question arises whether susceptibility to false memories and FIA effect go along with different personality characteristics. If it could be shown that susceptibility to false memories and FIA are unrelated phenomena that have their own distinct personality correlates, this would underpin the clinical impression that there are two different pathways to “recovering” memories. A number of studies have looked at whether certain personality traits contribute to susceptibility to false memories. Many of them examined the links between fantasy proneness (i.e., an extensive involvement in fantasy, vivid imagery, and daydreaming), dissociative symptoms, (e.g., derealisation, depersonalization, identity confusion), and false memories in undergraduate samples. For example, Jelicic and co-workers (2006) found a link between the creation of memories of a non-existent film and fantasy proneness as measured by the Creative Experiences Questionnaire (CEQ; Merckelbach, Horselenberg, & Muris, 2001). That is, participants who reported false details (i.e., memories of the non-existent film) scored significantly higher on the CEQ relative to those who did not report such details. Furthermore, positive associations between false memories and traits closely related to fantasy proneness (e.g., absorption and creative imagination) have been found with other paradigms (Drivdahl and Zaragoza, 2001 and Hyman and Billings, 1998). In addition to studies focussing on the link between false memories and fantasy proneness, researchers have attempted to relate false memories to dissociative symptoms. For example, Clancy and co-workers (2000) found a significant correlation between performance on the Dissociative Experiences Scale (DES questionnaire; Bernstein & Putnam, 1986) and DRM false recognition in people reporting continuous and recovered memories of CSA (Clancy et al., 2000). Also, Dehon, Bastin, and Larøi (2008) found that scores on the DES were positively associated with increased DRM false recall rates. In another study using a misinformation paradigm in undergraduate students, Hyman and Billings (1998) reported a positive association between false memories and dissociative experiences, and suggested that habitual dissociation may facilitate the acceptance of external information as self-defining, thereby increasing the likelihood to accept falsely suggested events as autobiographical memories. Using a variety of stimulus materials (e.g., video footage, slides), work of Giesbrecht and colleagues (Candel et al., 2003, Giesbrecht et al., 2007 and Merckelbach et al., 2007) also showed that there is a robust connection between dissociative symptoms and false memories (i.e., memory commission errors) in undergraduates. A link between false memory development and certain individual differences has also been documented in clinical samples. For example, Peters, Horselenberg, Jelicic, and Merckelbach (2007) showed that people with previous-life memories (i.e., memories of highly unlikely events) report elevated levels of dissociation. Like dissociation, poor cognitive efficiency (i.e., reporting many cognitive failures) has been associated with memory dysfunctions (Merckelbach, Muris, Nijman, & de Jong, 1996). Though research on individual differences and the FIA effect is rather scarce, one could speculate that people showing a large FIA effect will also score high on measures of cognitive failures (i.e., everyday lapses in attention, memory and perception, e.g., Broadbent, Cooper, Fitzgerald, & Parkes, 1982). One widely used measure of cognitive efficiency is the Cognitive Failures Questionnaire (CFQ; Broadbent et al., 1982) in which cognitive efficiency is approached in terms of susceptibility to cognitive failures. Wilhelm, Witthöft, and Schipolowski (2010) identified three factors underlying CFQ scores which were labeled retrieval (e.g., ‘Are you unable to remember something that you had been told some time ago?’), intention forgotten (e.g., ‘Do you find you forget what you came to the shops to buy?’), and clumsiness (e.g., ‘Do you bump into objects?’). Given that the overall content of the CFQ seems to have much in common with the type of failure that occurs in the FIA effect, one would predict they both covary. We investigated: (i) whether the FIA effect (i.e., the underestimation of prior remembering) and susceptibility to false memories are relatively independent from each other. We investigated this issue in an undergraduate sample (Study 1) and in a sample consisting of people with a history of childhood sexual abuse (Study 2). Moreover, we examined (ii) whether these memory illusions are related to different individual difference measures (Study 1). We hypothesized that the DRM and FIA effects would be unrelated to each other in the student as well as in the trauma sample. We also expected that these measures would be related to different individual difference measures. Specifically, we tested whether the susceptibility to false memories is particularly related to heightened levels of self-reported dissociation and fantasy proneness. On the other hand, susceptibility to underestimate prior memories was predicted to be particularly related to cognitive failures (Study 2). To avoid potentially confounding factors (e.g., depression, motivational variables), we started out with a non-clinical undergraduate sample to examine the relation between false memory propensity and the susceptibility to underestimate prior memories. One might argue, though, that results found in an undergraduate sample cannot be generalized to results in people with a trauma history. Particularly, differences in mean education level, mean age, and potential psychopathology between student samples and trauma samples could lower the generalizability of the findings. In Study 2, we therefore examined the relation between false memory propensity and susceptibility to the FIA effect in people with a history of childhood sexual abuse. To increase ecological validity, we used an autobiographical memory dating task which can be viewed as an autobiographical version of the FIA task described above (Parks, 1999).