This study tests the causal role of negative self-imagery in social anxiety. Low public-speaking anxious volunteers rehearsed a negative self-image, a positive self-image or a control image prior to giving a speech. As predicted, the negative image group felt more anxious, believed they performed less well and reported more negative thoughts than the positive image group. These findings do not appear to be due to changes in state anxiety, since they remained unchanged when anxiety was controlled in an analysis of covariance. The negative image group also reported more anxiety than the control group. Given that participants do not currently have anxiety problems, the findings are consistent with the idea that negative self-imagery has a causal role in the development and maintenance of social anxiety.
People with social phobias often report that, during social situations, they experience negatively valenced images of themselves (Hackmann, Clark, & McManus, 2000). The self-images typically represent exaggerated versions of themselves displaying socially inappropriate behaviour or observable anxiety symptoms, and may be based on memories of actual events dating from the onset of their social anxiety, although the individuals concerned are not always aware of this connection. The same or similar images tend to recur in anxiety-provoking social situations, and attention to such negative self-related representations, at the expense of external social cues, is thought to play a critical role in maintaining social phobia (Clark & Wells, 1995).
The proposition that negative self-images play a key role in the maintenance of social anxiety has been supported by the findings of recent experiments in which the type of image being held in mind was experimentally manipulated. In one such study (Hirsch, Clark, Mathews, & Williams, 2003), individuals with social phobia were required to converse with a stranger while holding in mind either their usual negative self-image, or a less negative control image. After the conversation in which the more negative image was being held in mind, participants rated their anxiety as greater, and their social performance as worse, than they did for the conversation when the control image was being held. Importantly, an observer who rated the videotape and was not informed about the type of image being held in mind, also rated the social performance of participants as being better in the control condition, than when they were holding the more negative image.
In a second study, Hirsch, Meynen, and Clark (2004) asked non-clinical participants, who were high in social anxiety, to hold two conversations with another volunteer whom they had not met previously. As before, during one conversation the socially anxious participants were asked to hold a negative self-image in mind, during another conversation they were asked to hold a control image in mind. As predicted, participants rated their anxiety as higher and believed they performed more poorly, when holding the more negative image. Ratings obtained from the volunteer with whom they conversed (who was not aware that the study related to imagery or social anxiety) confirmed that observers also perceived social performance, and the quality of the conversation overall, as being less good when the negative image was being held by the socially anxious participant.
Both the above studies thus suggest that the content of self-imagery can be manipulated experimentally, and that holding negative self-images in mind can cause both increases in anxiety and impoverished social performance. Within the theory of social phobia proposed by Clark and Wells (1995), self-images are thought to be based on memories of events that occurred around the time of onset of the disorder and are re-activated when entering social situations, thus causing anxiety to increase (Hackmann et al., 2000). A further test of the causal role of self-imagery would be to investigate if rehearsing a negative self-image just before entering a potentially threatening social situation, might have adverse consequences even for individuals who are not normally prone to social anxiety.
The present study attempts just such a test, by recruiting volunteers who reported relatively low levels of anxiety about speaking in public, and then requiring them to rehearse either a positive or negative self-image, developed by modifying an earlier memory of a related event. Participants were then required to give a brief speech, without being given any further instructions to hold the previously rehearsed image in mind. Our hypothesis was that anxiety would be greater and the perceived quality of performance would be poorer, in the negative relative to the positive image priming condition. Furthermore, we supposed that priming a negatively valenced emotional image might increase later cognitive intrusions, related to the content of the rehearsed image, which would be experienced while giving the speech. We also investigated if participants would perceive any actual consequences that were related to the content of these intrusions while they were speaking. A secondary question to be investigated was whether negative or positive images, or both, differed from baseline in a congruent direction. For this reason we included a control condition that manipulated imagery in a manner that was unrelated to public speaking. Finally, we assessed anxious mood immediately after image rehearsal, to investigate if any effects could be attributed to the direct consequences of changes in mood prior to giving the speech, as opposed to delayed priming effects of the imagery manipulation.
