تفکیک، خطاهای کمیسیون حافظه، و واکنش پذیری افزایش اتونوم

Abstract If dissociative symptoms are manifestations of a psychological defense mechanism, one would expect people who have such symptoms to react with lower levels of physiological arousal and with memory omissions to emotionally provocative material. The current study tested this assumption in a sample of undergraduates. Sixty-two undergraduate students viewed a highly emotional video fragment. Pearson's product–moment correlations were calculated between dissociation (as indexed by the Dissociative Experiences Scale, DES) and all dependent measures. High dissociators exhibited elevated skin conductance responses (SCRs) to the fragment. Memory for the video fragment was then tested. While omission errors were unrelated to dissociation, high dissociators exhibited a tendency to produce commission errors. This could not be explained by a reduced working memory capacity. However, fantasy proneness was found related to high dissociators' commission errors. Thus, it seems that a pattern of heightened emotional reactivity and commission errors is typical for people with elevated dissociation scores. This pattern is difficult to reconcile with the defensive function ascribed to dissociation.

1. Introduction According to Bernstein and Putnam (1986, p. 727), dissociation is “a lack of normal integration of thoughts, feelings, and experiences into the stream of consciousness and memory”. Many authors believe that it “functions to fragmentize, derealize, and depersonalize traumatic experiences” (Markowitsch, 2003, p. S133). Dissociation can be conceptualized as forming a continuum that ranges from the minor dissociation of everyday life to major forms of psychopathology (Fischer and Elnitsky, 1990). Historically, dissociation has been considered a defense mechanism that enables the individual to withdraw psychologically from the impact of overwhelming traumatic events (Gershuny and Thayer, 1999). This view dates back to the 19th century psychologist Pierre Janet (1894), who coined the term “dissociation" and emphasized its role as a defensive maneuver in response to psychological trauma. A habitual tendency to dissociate would, however, promote psychopathology (Hacking, 1995). Following this tradition, recent clinical literature emphasizes the causal relationship between dissociation and trauma (Irvin, 1998, Gershuny and Thayer, 1999 and Gast et al., 2001). While the view that dissociation is caused by exposure to traumatic events is widely accepted, empirical studies providing unequivocal support for this causal direction are still lacking. For example, Zlotnick et al. (1996) found a correlation of 0.40 between dissociation, as measured with the Dissociative Experiences Scale, and their own Sexual Assault Questionnaire (SAQ). Yet, the authors admit that the validity and the reliability of the SAQ are unknown. Similarly, Nash et al. (1993) investigated the link between trauma and dissociation in a sample of 105 women. They found that participants with a self-reported history of sexual abuse endorsed significantly more dissociative symptoms compared with women without such a history. However, this connection disappeared when a measure of family pathology was entered as a covariate in the analysis. This suggests that the connection between trauma and dissociation might be less straightforward than is often thought. This point is emphasized by Mulder et al. (1998), who summarized their findings in a large community study (n = 1000) as follows: Any causal influence of childhood sexual abuse on dissociation is likely to be indirect and mediated by more general linkages between childhood sexual abuse and risks of mental disorder. (p. 809) Another study yielding problematic results for the trauma-dissociation hypothesis is that by Sanders and Giolas (1991). These authors found retrospective self-reports of traumatic experiences and dissociative symptoms to be modestly correlated in a sample of adolescent psychiatric patients. Curiously, when hospital records were scored by a “blind” rater for indications of trauma, a slightly negative correlation emerged between ratings of traumatic experiences based on hospital records and dissociation. This suggests that dissociation is related to retrospective self-reports of traumatic experiences, while being unrelated to more objective and specific indices of such experiences. Recently, Merckelbach and Jelicic (2004) offered an explanation for this phenomenon. In two samples of undergraduate students (n = 43 and n = 127), they found that dissociative experiences are related to the endorsement of vague, but not specific trauma items. This response tendency might also have played a role in other studies that found a link between dissociation and trauma as measured by retrospective self-report scales of traumatic experiences. Mild dissociative symptoms are rather common in the general population, with 80% to 90% of the respondents indicating that they sometimes experience these symptoms (Gershuny and Thayer, 1999). This is difficult to reconcile with the alleged traumatic etiology of dissociative symptoms, which implies that they should have a lower base rate. In addition, Lang et al. (1998) investigated the influence of heritability on dissociative experiences using a twin study. They report that about half of the variability in dissociative experiences can be attributed to heritability. The evidence concerning the genetic predisposition of dissociative experiences is, however, mixed, as another study (Waller and Ross, 1997) reports no genetic influence, while a recent study in children and adolescents lends support to the notion of a genetic predisposition of dissociation (Becker-Blease et al., 2004). Thus, the high prevalence and the possible heritable basis of dissociative phenomena make a direct causal link between dissociation and trauma less likely. It also underlines the importance of a different research perspective, namely one that focuses on the trait-like features of dissociation. One robust finding in this domain is that there exists a substantial overlap between dissociation and fantasy proneness (Rauschenberger and Lynn, 1995, Merckelbach et al., 2000 and Waldo and Merritt, 2000). Fantasy proneness refers to a deep and extensive involvement in fantasizing and daydreaming (Merckelbach and Muris, 2001). While fantasy proneness is integral to healthy psychological functioning, an inability to control the cognitive processes implicated in imaginative involvement may be associated with psychopathology (Rauschenberger and Lynn, 1995). Fantasy proneness has repeatedly been shown to be related to dissociation in both clinical (Pekala et al., 1999–2000 and Merckelbach et al., 2005) and non-clinical samples (Merckelbach et al., 1999 and Merckelbach et al., 2000) with correlations ranging from 0.41 (Pekala et al., 1999–2000) to 0.55 (Merckelbach et al., 2000). Given the fact that fantasy proneness is associated with susceptibility to pseudo-memories, dissociation might also be related to a tendency to confabulate. Evidence for this comes from a study by Merckelbach et al. (2000). Participants watched 40 slides, with 20 depicting photographs of common objects or situations and the other 20 being a short paragraph describing a scene or situation. Following this, participants were given a surprise recognition test in which they had to identify slides they had seen earlier in a series of old and new slides. They also had to indicate whether they had seen the “old” slide as a photograph or a paragraph. Two types of errors may occur during this task. Participants may erroneously confuse a picture with a text or vice versa. This would be a reality-monitoring error. Participants might also claim that they recognize a slide, which was not presented earlier. This would be a pseudo-memory, i.e., a commission error. Dissociative experiences were found to be related to commission errors, but not to reality-monitoring errors. In a more recent study, Candel et al. (2003) examined memory-commission errors in a sample of undergraduate students scoring either high (n = 19) or low (n = 19) on the DES. The authors presented an emotional narrative to the participants, who then had to complete a free-recall task. Dissociation was associated with commission errors, but not with the proportion of correct recall, with high dissociators committing significantly more commission errors than low dissociators. In contrast to other studies ( Merckelbach et al., 2000), the link between dissociation and commission errors was not mediated by fantasy proneness. This, however, might have been due to the limited sample size in the Candel et al. (2003) study. Our study attempted to replicate the findings of Candel et al. (2003), but also tried to extend our understanding as to why commission errors are associated with heightened dissociation. One possibility is that this association has to do with high dissociators suffering from limited working memory capacity. A limited working memory capacity might hamper encoding, leading to gaps in memory (Brewin and Saunders, 2001), which are later filled in by participants, thereby leading to confabulations. Thus, the first issue examined in the current study is whether high dissociators' commission errors are related to working memory problems. A second issue addressed by our study has to do with the defensive function that is often ascribed to dissociative experiences (Gershuny and Thayer, 1999). If this idea is correct, one would hypothesize that high dissociators show fewer hits on a memory task for emotional material (i.e., one would expect more omission errors), as one of the core features of dissociation is psychogenic amnesia (American Psychiatric Association, 1994). During exposure to the emotion-provoking material, one would also expect them to exhibit attenuated skin conductance responses as compared with low dissociators (e.g., Simeon, 2004). That is, one would predict that high dissociators could withdraw themselves psychologically from emotional events. Alternatively, if dissociative experiences and fantasy proneness described a common domain, high dissociators' skin conductance responses would be more pronounced as compared with those of low dissociators due to their more pronounced imaginative involvement in the emotional material (Candel and Merckelbach, 2003).

3. Results 3.1. Psychometric instruments 3.1.1. Individual differences measures Mean DES and CEQ scores were 17.86 (S.D. = 10.46) and 7.22 (S.D. = 3.73), respectively. These scores correspond with values that previous studies reported for student samples (Merckelbach et al., 2002). Table 1 presents Pearson's product–moment correlations between the DES, the CEQ, and performance on the digit span task. Table 1. Pearson's product–moment correlations between dissociative experiences (DES), fantasy proneness (CEQ), digit span, free-recall performance, and commission errors for an undergraduate sample (n = 62) Individual difference measures DES CEQ Individual difference measures DES – 0.51* CEQ 0.51* – Digit span Forward − 0.11 0.10 Backward 0.15 0.11 Free recall Proportion hits − 0.07 − 0.06 Commission errors 0.34* 0.36* Only relevant correlations are shown. * Correlation is significant at the 0.01 level (2-tailed). Table options As can be seen, there was a significant correlation between DES and CEQ scores. Thus, participants who reported more dissociative experiences exhibited stronger tendencies to fantasize. 3.1.2. Digit Span Digit span forward and backward scores were 6.18 (S.D. = 1.11) and 4.65 (S.D. = 1.10), respectively. Correlations between digit span forward, backward, and dissociative experiences remained non-significant. 3.2. Recall performance 3.2.1. Free recall Inter-rater reliability values for hits and commissions were 0.85 and 0.82, respectively. Participants recalled on average 36% (S.D. = 8%) of the critical details, while making on average 0.37 (S.D. = 0.58) commission errors (in absolute numbers). Pearson's product–moment correlations showed that both individual differences in dissociative experiences and fantasy proneness were positively related to commission errors, while not affecting hits (see Table 1). It is worth noting that previous research (Smeets et al., 2004) in our laboratory indicates that recall or emotional impact is unaffected by whether participants are familiar with the video fragment. One might wonder whether the finding that dissociation and fantasy proneness are related to the number of commission errors in memory is confounded by the self-paced nature of our free-recall task. The length of a written report might exert a strong influence on the reported content. That is why we calculated Pearson's correlations between the number of words in the written report and the number of hits and commission errors. As expected, the number of hits was positively related to the number of words (r = 0.64, P < 0.01); however, more interestingly, commission errors were unrelated to the number of words (r = − 0.03, P > 0.05). Therefore, our finding that heightened levels of dissociation and fantasy proneness were related to commission errors in memory is not dependent on the length of the report participants give. 3.2.2. Modeling fantasy proneness as mediator during free recall The mediational effect of fantasy proneness on the relationship between dissociation and commission errors during the free-recall task was examined following the approach recommended by Baron and Kenny (1986). Firstly, we tested the unmediated model, consisting of the link between dissociation and commission errors in memory without controlling for fantasy proneness. Dissociative experiences, as measured by the DES, were found to predict commission errors during the free-recall task, B = 0.02, β = 0.34, t(61) = 2.78, P < 0.01. Next, the mediated model was tested. This model consisted of dissociative tendencies, fantasy proneness, and commission errors during the free-recall task. Dissociative tendencies predicted fantasy proneness, B = 0.18, β = 0.51, t(61) = 4.60, P < 0.01, whereas a trend emerged for fantasy proneness to predict a tendency to commit memory commission errors, B = 0.04, β = 0.25, t(61) = 21.14, P = 0.07. In addition, the effect of dissociation on memory commission errors decreased and was no longer significant, B = 0.01, β = 0.21, t(61) = 1.51, P = 0.14. However, as it is not sufficient to show that the link between the independent variable and the dependent variable becomes non-significant when the mediator is added to the model ( Frazier et al., 2004), we calculated Sobel's (1982) test. This test indicates whether the indirect influence of the independent variable via the mediator is different from zero. Only a trend emerged, z = 1.69, P = 0.09, with 38% of the relation between dissociation and commission errors in free-recall task being mediated by fantasy proneness. 3.3. Skin conductance Dissociative experiences were positively correlated with SCR magnitude: r = 0.34, P = 0.01, whereas correlations between dissociative experiences and SC baseline level remained non-significant (r = − 0.15, P > 0.10). Fantasy proneness was not related to SCR magnitude (r = 0.18, P > 0.10). Dissociation was unrelated to the SC decrement following the maximum SCR (r = 0.03, P > 0.10). While there is no reason to expect that the 4-s window described above should differentially affect high and low dissociators, we excluded this possibility by calculating the maximum SCR during the whole video fragment. This approach did not alter the correlation between dissociation and the SCR (r = 0.34, P < 0.01).