آلکسیتیمیا ناهماهنگی بین اعتماد به نفس ضمنی و صریح را افزایش می دهد

Recent studies have stressed the importance of affective information in the translation of implicit/associative evaluations into an explicit/propositional format. Accordingly, we predicted that alexithymia (as an inability to process emotions explicitly) increases implicit vs. explicit self-esteem (SE) discordance. Subjects were 310 university students with mean age 20.5. Four multiple regression analyses with interaction were conducted on self-reported measures of SE, using SE-IAT (Self-Esteem Implicit Association Test) and TAS-20 (Toronto Alexithymia Scale) scale and subscales scores as predictors (Alexithymia total score; Difficulty Identifying Feelings – DIF; Difficulty Describing Feelings – DDF and Externally Oriented Thinking – EOT). In the first regression the interaction term was significant and in the expected direction, confirming that the alexithymia total score increases implicit–explicit SE discordance. With the TAS-20 subscales, only DIF and DDF, but not EOT showed the expected impact on implicit–explicit SE consistency. Some implications in a clinical perspective are discussed.

In the last decade a number of researchers have begun to use indirect testing procedures to measure relevant psychological constructs like attitudes, stereotypes, self-esteem and personality traits. One of the most popular of these procedures is the Implicit Association Test (IAT), first developed by Greenwald, McGhee, and Schwartz (1998) and successively used in many areas, among them in the empirical assessment of self-esteem and self-concept (Greenwald & Farnham, 2000). The present study was aimed at demonstrating that the relation between implicit and explicit measures of self-esteem is moderated by the ability to elaborate feelings at the explicit level. Concerning the relation between implicit and explicit measures, recent meta-analyses covering a series of content domains, including self-esteem and self-concept, showed average implicit–explicit correlations of moderate size: .24 (Hofmann, Gawronski, Gschwendner, Le, & Schmitt, 2005), .37 (Nosek, 2005) and .21 (Greenwald, Poehlman, Uhlmann, & Banaji, 2009). These correlations varied considerably among content domains, with aggregate values as large as .54 for political preference and as small as .09 and .12 for close relationships and racial prejudice ( Greenwald et al., 2009). In a paper on the theoretical distinction between implicit and explicit constructs, Nosek and Smyth (2007) used a multitrait–multimethod approach across seven-attitude objects (flower–insect, creation–evolution, democrat–republican, humanities–science, straight–gay, thin–fat and white–black) to demonstrate that implicit and explicit measures refer to distinct, though related, constructs. In particular, they searched for the best fitting solution among three models: a model with two method factors (whether implicit or explicit measures were used), a model with two method factors plus a single attitude factor for each object (that did not distinguish an explicit from an implicit attitude, e.g. a single unitary attitude towards democrat–republican, etc.), and a model with two method factors plus two correlated attitude factors for each object (e.g. an implicit attitude and an explicit attitude towards democrat–republican, etc.). The last solution obtained the best fit. In this model each IAT correlated with the explicit attitude measures toward the same object (and not with explicit measures toward other objects), but both IAT and the explicit measures also retained unique components that were not reducible to shared method variance. Along similar lines, Greenwald and Nosek (2008) reported the results of a study on over 10,000 subjects showing apparent dissociations in the form of (a) weak correlations between implicit and explicit measures, (b) separation of their means on scales that should have coincided if they assessed the same construct and (c) differing correlations with other criterion variables. From an analysis of the relevant literature they identified three main interpretations of the implicit–explicit relation: single representation – the two types of measures assess a single attitude, but under the influence of different extra-attitudinal processes ( Fazio & Olson, 2003); dual representation – the two types of measures assess distinct forms of attitudes ( Chaiken et al., 1999 and Strack and Deutsch, 2004); person vs. culture – a variant of the dual representation view, in which self-report measures reflect personal attitudes, whereas IAT measures reflect non-attitudinal cultural semantic knowledge ( Olson & Fazio, 2004); and concluded that, although the empirical evidence supported the distinction between implicit and explicit constructs, it could not indicate which model of representation (single or dual) was better. That is, the distinction between implicit and explicit constructs may be interpreted in terms of either one or two representations and the choice between the two alternatives should only be based on parsimony and explanatory power. As regards dual representation theories, Gawronski and Bodenhausen (2006) proposed the APE model (APE: associative and propositional evaluation), positing that indirectly assessed attitudes reflect associative evaluations toward an attitude-object, whereas self-reported attitudes reflect propositional evaluations. For APE the two kinds of evaluations do not only differ for their conscious/unconscious level, but also for their processing modes. Associative evaluations depend on links between conceptual elements that are activated spontaneously in memory, producing immediate affective reactions toward the attitude-object. These conceptual links are formed according to the principles of contiguity and similarity, and no truth values are assigned to them. In contrast, propositional evaluations reflect evaluative judgements toward an attitude object, that is judgements that are founded upon logical rules involving the assignment of truth values, that may or may not be based on immediate affective reactions. If the affective reaction is consistent with other information that is used for the propositional evaluation, then the explicit attitude usually reflects the evaluative quality of the implicit attitude. However, if the affective reaction is inconsistent with the propositional information activated, this reaction may be rejected as a valid basis for an evaluative judgement. In line with the APE model, Hofmann, Gschwendner, Nosek and Schmitt (2005) proposed a working model of implicit–explicit consistency that organizes the empirical evidence within five general groups of moderating factors. Of interest here is the moderating role attributed by them to the factors pertaining to awareness. They argued that the translation process from an implicit to an explicit format may depend on the degree to which people are able to form accurate propositional representations of their underlying associative representations. We may become aware of our implicit associative evaluations in two ways: we may be aware of the internal outcomes (i.e., our gutfeelings) of our implicit associations (e.g. between black men and bad values), or we could generate inferences about our implicit evaluations through the observation of our own behaviour ( Bem, 1972). To support the moderating role of the awareness factors, Hofmann, Gschwendner, et al. (2005) reported the results of studies that examined the influence of private self-consciousness and mindfulness on implicit–explicit consistency. Private self-consciousness was defined as the extent to which individuals pay attention to their bodily and emotional experience and mindfulness as an enhanced attention to and awareness of current experience or present reality, which may be reflected in more regular or sustained consciousness of ongoing events and experiences. The hypothesis was that high scorers in private self-consciousness and in mindfulness exhibit stronger implicit–explicit consistency than low scorers. The effect of private self-consciousness was found to be significant only in one out of three studies ( Gschwendner T., 2004 and Hofmann et al., 2005c), while in the only study devoted to mindfulness ( Brown & Ryan, 2003) high scorers in this trait showed stronger concordance between implicit and explicit assessment of affective states than low scorers. More recently, Smith and Nosek (unpublished manuscript), adopted an experimental approach, based on manipulation, rather than a correlational approach and demonstrated that focusing attention on affective information (i.e., feelings and emotions) rather than on cognitions (i.e., thoughts and beliefs) increased the concordance between implicit and explicit attitudes toward cats/dogs and gay/straight people. They concluded that explicit evaluations can be meaningfully parsed into affective and cognitive components, whereas implicit evaluations are related to affect rather than to cognition. With regard to self-esteem, Jordan, Whitfield and Zeigler-Hill (2007) argued that implicit self-esteem (ISE) is based on intuitive self-views that are connected to immediate affective reactions toward the self, whereas explicit self-esteem (ESE) is based on deliberative self-views that are connected to evaluative judgements. By manipulating the perceived validity of intuition, they demonstrated that subjects led to consider their intuition as more valid, reported ESE scores that corresponded more closely to ISE. More recently, Grumm, Nestler and Von Collani (2009 tested predictions concerning asymmetrical patterns of implicit and explicit self-esteem change. In particular, they demonstrated that evaluative conditioning influences implicit self-esteem but not explicit self-esteem, that manipulating the salience of the self-knowledge influences explicit but not implicit self-esteem and, finally, that evalutive conditioning can “spill over to the explicit level” when participants are asked to focus on their feelings prior to making their self-report judgements. Coming to the present study, if we accept that affective awareness has an influence on implicit–explicit consistency, we may predict that an impaired capacity to identify feelings and to express them through verbal language as reflected by individual differences of alexithymia (Sifneos, 1972 and Lane and Schwartz, 1987), will tend to produce a lack of concordance between implicit and explicit measures of attitudes and personality related variables including self-esteem. In other terms, our hypothesis is that alexithymic individuals are impaired in the capacity to explicitly elaborate implicit feelings, and hence their explicit self-judgements will be less informed by implicit feelings.

The study showed that alexithymia (as measured by TAS-20 total score) moderates the concordance between implicit and explicit self-esteem. As an illustration, it was found that at one standard deviation below the mean of the TAS-20 scale there was a positive, though moderate, correlation between implicit and explicit measures of self-esteem, whereas at one standard deviation above the mean the correlation was not significant and close to zero. The interaction between alexithymia and implicit self-esteem was also found for the subscales of Difficulty Identifying Feelings and Difficulty Describing Feelings, but not for Externally Oriented Thinking, and it appeared to be due to what the two former subscales have in common. The lack of interaction between EOT and implicit self-esteem may be explained by noting that, compared to the other alexithymia subscales, EOT seems to be less directly related to the awareness of feelings. However another, not necessarily alternative, explanation may be given in terms of the lower reliability of this subscale compared to the other subscales. Generally, these results give further support to the role ascribed to emotional awareness in the propositional/explicit translation of affective/implicit evaluations (Hofmann, Gschwendner, et al., 2005, Jordan et al., 2007 and Smith and Nosek, 2010 and are compatible with Hofmann, Gschwendner, et al. (2005) working model of implicit–explicit consistency, that posits awareness in a general group of moderators called translation. These findings may have interesting clinical implications. Our data may indicate that alexithymia may hamper the explicit monitoring of affective (implicit) self-evaluations that are likely to fluctuate according to life events and situations – fluctuations that must be translated into a propositional format to be fully perceived at a conscious level. As a result, alexithymics may be unable to identify and cope with the stressors that threaten their self-image and may thus develop habits that are dysfunctional for their self-esteem. For example they may tend to expose themselves to pathogenic situations that threaten their self-esteem, recognizing them when it is too late. In other words, given that an accurate translation of SE gut feelings into propositional format may help to cope with potentially pathogenic situations, low alexithymic subjects should be less prone to developing mental suffering than high alexithymic subjects. Along similar lines, Schröder-Abé, Rudolph and Schüz (2007) showed that a high discordance between implicit and explicit SE is related to important indexes of mental and physical health, such as anger suppression, depressive attributional style, nervousness, and days of impaired health. In their view implicit–explicit SE discordance determines a weakness in self-integrity that induces unpleasant emotions and internal conflicts, directly connected to proneness toward mental diseases. We agree with Schröder-Abe et al. but add that an inability to cope with self-threatening situations, brought about by a defective translation of affective information into propositional format, may, on its own, be another source of mental suffering. Clearly, empirical studies on clinical samples are needed to identify the specific impacts of these factors. A limit of the present study is that alexithymia was assessed only via a self-report inventory. An important confirmation of our results may come from investigations using clinical assessments of alexithymia, e.g. via the Toronto Structured Interview for Alexithymia ( Bagby, Taylor, Parker, & Dickens, 2006), or directly manipulating it as an experimental variable. Another limitation of the present study is that implicit self-esteem was measured only through classic IAT. It may be appropriate for future research to check the robustness of our results by also using other implicit techniques such as the Go/No-Go Association Task (GNAT; Nosek & Banaji, 2001) or the Single Target Implicit Association Test (ST-IAT; Bluemke & Friese, 2008) that may also use one target category rather than two.