دیدگاه های ضمنی از خود در اضطراب اجتماعی

We investigated dysfunctional attitudes in high (N=29) and low (N=28) socially anxious participants following a social-threat activation task (being told to give a speech) using the implicit association task (IAT: [Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The implicit association task. Journal of Personality and Social Psychology, 74, 1464–1480]) and the Social Attitudes Questionnaire (SAQ: [Clark, D. M. (2001). Social Attitudes Questionnaire, revised. Unpublished manuscript, University of Oxford]). The study also looked at anticipatory thoughts about the speech using questionnaire (Social Cognitions Questionnaire (SCQ): [Wells, Stopa, & Clark (1993). The social cognitions questionnaire. Unpublished]) and think aloud methods. High socially anxious participants endorsed more negative attitudes on the SAQ. In the IAT both groups showed a bias towards associating positive words with self and negative words with other but this was weaker in the high socially anxious group than in the low. The high socially anxious group endorsed more negative thoughts about the speech on the SCQ, believed these thoughts more, and had more thoughts overall on the think-aloud task. There were no differences between the groups on the think-aloud task, but there was an interaction between level of depressive symptoms and thought valence. The results are considered in relation to other studies that have used an IAT to measure associations in social anxiety and in relation to Clark and Wells’ [(1995). A cognitive model of social phobia. In R. G. Heimberg, M. Liebowitz, D. A. Hope, & F. Schneier (Eds.), Social phobia: Diagnosis, assessment and treatment. New York: Guilford] model of social phobia.

Clark and Wells’ (1995) cognitive model of social phobia proposes that individuals with social phobia interpret social situations as threatening because they hold dysfunctional assumptions about the nature and consequences of social performance and about the acceptability or worth of the self. They identified three categories of assumptions: excessively high standards for social performance, conditional beliefs concerning social evaluation, and unconditional beliefs about the self. In the model, assumptions (if…then rules) and unconditional beliefs (I am…) are activated by social situations. Clark and Wells argue that in comparison to depression, where unconditional beliefs are global and stable, beliefs in social phobia are unstable and are activated by anticipating or participating in social interactions. However, social phobia has a high rate of co-morbidity with depression (Schneier, Johnson, Hornig, Liebowitz, & Weissman, 1992), which may lead to a more stable negative self-view than when social phobia presents alone. This would support Clark and Wells’ proposal that there is considerable variance between individuals in the degree to which all three types of dysfunctional beliefs are held. They also suggest that individuals with generalised social phobia, who have developed social anxiety at an early age, are more likely to hold negative unconditional beliefs about the self. Despite the importance of assumptions in the cognitive model of social phobia, there is relatively little empirical evidence in support of their role in social anxiety. Investigation of assumptions in other disorders such as depression has relied heavily on self-report measures such as the dysfunctional attitudes scale (DAS: Weissman & Beck, 1978). However, self-report is problematic because it can be influenced by response bias and demand characteristics (Nisbett & Wilson, 1977). Self-report also requires appropriate and specific questionnaires, individuals must be able and willing to report their cognitions accurately, and the measured beliefs must be available to consciousness (for example, Segal & Swallow (1994) argue that schema level processes operate in a non-conscious manner). In this study we combined a self-report measure designed specifically to look at assumptions in social phobia with two versions of the Implicit Association Test (IAT) developed for this experiment. The IAT was originally developed by Greenwald, McGhee, and Schwartz (1998) to investigate implicit associations between two concepts. Greenwald et al. asked participants to classify the names of flowers (e.g. TULIP) and insects (e.g. SPIDER) and positive (e.g. LOVE) and negative words (e.g. UGLY) by pressing one of two keys. The idea behind the IAT is that participants will be faster to respond when compatible concepts share a response key (e.g. flower+positive and insect+negative) because these two concepts are associated in memory, than when incompatible concepts share a response key (e.g. flower+negative and insect+positive). The IAT has been adapted by clinical researchers to investigate beliefs in different types of psychopathology (see De Houwer, 2002, for a review). De Houwer, points out that the IAT measures associations rather than beliefs. However, he argues that associations are likely to be part of beliefs, and therefore the IAT can provide an indirect index of the existence of dysfunctional beliefs that is free from many of the demand characteristics that limit self-report measures. Two recent studies have used the IAT to investigate beliefs in social anxiety. In the first, de Jong, Pasman, Kindt, and van den Hout (2001) looked at whether the IAT could differentiate between high and low socially anxious women. de Jong et al. used the IAT to measure an association between a single target concept (SOCIAL) and a control category (words describing rooms in a house) plus positive and negative words that related to social outcomes (e.g. success and failure). de Jong et al. argued that for low socially anxious individuals, social words (e.g. DATE) and positive words (e.g. SUCCESS) would be compatible, but for high socially anxious participants the opposite would be the case. In their analysis of reaction time data, de Jong et al. found a trend in support of their hypothesis that high socially anxious women would be faster when social and negative words were assigned to the same key (incompatible task) and that low socially anxious women would be faster when social and positive words were assigned to the same key (compatible task). When they analysed error rates, there was clear evidence that the high socially anxious group made more errors in the compatible task (social+positive), whereas low socially anxious participants did the opposite. However, De Houwer (2002) pointed out that this approach—using a single target concept—has not been used by other IAT researchers and is not validated. In the second study, de Jong (2002) used a more conventional IAT to examine self- and other-esteem and found that although both high and low socially anxious participants showed a bias towards associating positive words with self, this bias was weaker in the high socially anxious group. Reduced levels of self-esteem have previously been associated with depressive symptoms (Vazquez, 1987), and the high socially anxious group in de Jong's study may have had higher depression scores than the low socially anxious group. Neither of the two IAT studies described above (de Jong, 2002; de Jong et al., 2001) measured depressive symptoms. A further criticism of the two studies is that neither of them made any attempt to activate dysfunctional attitudes. Clark and Wells (1995) argued that dysfunctional assumptions are activated by social events and the absence of a social threat activation task might mean that these studies did not demonstrate the full extent of negative attitudes towards the self. The main aim of the present study was to investigate implicit attitudes to self using an IAT following a social threat activation (giving a short speech). The original study had planned to compare attitudes before and after the social threat priming task but data from the first task was lost due to an administrative error while running the study and therefore only the results of the IAT that was completed after the social threat activation task can be reported. In the IAT, self-items, other items, positive words, and negative words were presented. The positive and negative words were selected on the basis of how well they represented a positive or negative assumption about the self in social anxiety. The study also had two subsidiary aims. The first was to examine dysfunctional assumptions using a self-report questionnaire that had been designed specifically for social anxiety (Social Attitudes Questionnaire (SAQ): Clark, 2001). The second was to examine differences between the groups in anticipatory thoughts about the speech. In previous studies examining self-esteem, participants were usually faster to respond in the compatible condition (self+positive and other+negative) than in the incompatible condition (self+negative and other+positive). We hypothesised that the low socially anxious group would be faster in the compatible condition, but that the high socially anxious group would be faster in the incompatible condition because they would be more likely to associate self-words with negative attributes and words denoting other people with positive attributes. We also hypothesised that high socially anxious participants would endorse more negative thoughts on the SAQ and that they would report more negative thoughts while anticipating giving the speech.