Images of the self in social anxiety: Effects on the retrieval of autobiographical memories

https://doi.org/10.1016/j.jbtep.2007.08.006Get rights and content

Abstract

Cognitive models of social phobia propose that negative self-images play an important role in maintaining anxiety. This study examines the effect of holding a positive or a negative image in mind during a speech on the retrieval of autobiographical memories. Twenty high socially anxious participants performed a standard autobiographical memory task (AMT), which used positive, negative and neutral cue words. Participants performed the AMT twice: once after giving a speech holding a positive image and once while holding a negative image. Participants were more anxious and rated their performance worse in the negative image condition. Negative memories were retrieved faster in the negative image condition and positive memories were retrieved faster in the positive image condition. In the negative image condition, positive memories were retrieved more slowly than either negative or neutral memories. Inhibition and facilitation are proposed as two processes that could explain the effects of differently valenced imagery on autobiographical memory. The clear evidence for an inhibitory effect on positive autobiographical memories in the negative imagery condition is considered in relation to Brewin's [(2006). Understanding cognitive behaviour therapy: A retrieval competition account. Behaviour Research and Therapy, 44, 765–784] retrieval competition hypothesis. The paper concludes with a brief discussion of the potential role of inhibition in imagery rescripting.

Introduction

In the current cognitive models of social phobia, images of the self play an important role in maintaining social anxiety (Clark & Wells, 1995; Rapee & Heimberg, 1997) and recent evidence suggests that these images could also play a causal role in the development of social anxiety (Hirsch, Matthews, Clark, Williams, & Morrison, 2006). The self-images of socially phobic individuals are frequently negative and distorted, and yet the individuals believe that they are accurate representations. The effects of this mental bias are pervasive and socially phobic individuals rate their appearance and attractiveness worse compared to observers, as well as underestimating their social and performance skills (e.g. see Rapee & Abbott, 2006).

Preliminary evidence (Wild, Hackmann, & Clark (in press a), Wild, Hackmann, & Clark (in press b)) suggests that imagery rescripting is a useful technique in therapy for social phobia. Wild et al. (in press b), for example, used a single session of imagery rescripting with a group of 20 patients with social phobia and demonstrated that rescripting can significantly decrease distress in response to the image, modify the beliefs about the self encapsulated by the image and reduce anxiety. Imagery rescripting involves modifying the meaning of an arousing or distressing image through the use of imagery techniques. This procedure, which targets the content of an image, aims to change the idiosyncratic meaning of the image to the individual and therefore reduces its power to elicit negative emotions. However, it is not clear exactly how imagery rescripting works.

Images of the self are constructed representations that are probably formed from multiple inputs. Rapee and Heimberg (1997), for example, argue that mental representations of the self draw on information retrieved from long-term memory as well as incorporating information from more immediate sources such as proprioceptive and physiological feedback. Clark and Wells (1995) point out that socially phobic individuals’ images of themselves often contain visible exaggerations, so that a woman who worries about blushing might see herself in the image as a vibrant fire-engine red.

Images are not only important during social interactions, they can also play a role in anticipatory and post-event processing, both of which contribute to the maintenance of social phobia. Anticipatory and post-event processing are forms of ruminative processing, in which the individual retrieves memories of either recent or more distant social experiences and dwells on the negative aspects of them. One possible difference between rumination in depression and in social anxiety could be the presence of imagery; for example, James (2005) found that high socially anxious participants had more negatively valenced images during post-event processing than low socially anxious individuals. Self-imagery can lead to increased perception of body sensations and worse ratings of performance (Vassilopoulos, 2005). Spontaneously recurring images in socially phobic individuals are often related to early memories of aversive social experiences (Hackmann, Clark, & McManus, 2000).

The constructed images of the self reported by the socially phobic individuals could be conceptualised as the representations of the individual's current working self. The working self refers to the subset of self-schemas that is active at any given time (Markus & Nurius, 1986). According to Conway and Pleydell-Pearce (2000), self and memory are inextricably linked in a self–memory system where the working self draws on autobiographical memories to both set and constrain current goals. If this is the case, then negative self-imagery may facilitate the retrieval of negative autobiographical memories. Alternatively, the retrieval of negative memories may influence the type of working self that is constructed or accessed in social situations, and when anticipating or reflecting on social interaction.

To date, the evidence for memory biases in social phobia has been weak (Rinck & Becker, 2005). However, autobiographical memory has only recently become a focus of study in social phobia. Wenzel, Jackson, and Holt (2002), for example, asked 16 socially phobic individuals and 17 non-anxious controls to write down memories in response to social-threat and neutral words. These memories were coded according to whether they were general (related to a recurring class of events or taking place over more than 1 day) or specific (event occurred at a particular time and place). Coders also rated the affective tone of the memories (positive, negative and neutral). There were no differences between the groups in the number of memories retrieved for each cue type. However, analysis of the affective tone showed that the socially phobic group reported a larger proportion of memories associated with negative affect in response to social-threat cue words compared to the control group. The absolute percentages of negative memories were very small (5% for the socially phobic group versus 3% for controls) and as a result Wenzel et al. concluded that the findings had limited practical significance.

Recent studies have been more promising. Field and Morgan (2004) investigated whether different types of post-event processing affect the retrieval of autobiographical memories. There was no effect of type of post-event processing on retrieval, but they did find that, after negative post-event processing, the socially anxious individuals remembered more memories that were rated as more calming, even though they were also rated as anxious and shameful. Wenzel and Cochran (2006) looked at the retrieval of autobiographical memories in response to panic-related, social-phobia related and control automatic thoughts in patients with anxiety disorders (panic disorder n=20, social phobia n=22) and non-anxious controls (n=20). Both the patient groups retrieved more anxious and worried memories in response to cues that were specific to their diagnoses; however, the patients also retrieved more fearful memories overall compared to the non-anxious control group, irrespective of type of cue. D’Argembeau, Van der Linden, d’Acremont, and Mayers (2006) examined the subjective distress associated with memories for social and non-social events in the socially phobic and non-anxious individuals. The socially phobic group recalled fewer sensory details but more self-referent information for social events compared to the non-anxious group. They were also more likely to remember the events from an observer perspective. The two groups did not differ in their recall of non-social events.

The preliminary evidence for the roles of both imagery and autobiographical memory in social phobia is promising. However, Hirsch, Clark, and Matthews (2006) argue persuasively that cognitive biases influence each other and that it is important to study these interactions. Their article focuses on the reciprocal relationship between self-imagery and interpretations in social phobia and demonstrates that negative imagery can interfere with the formation of benign inferences (e.g. Hirsch, Matthews et al., 2006); whereas positive imagery can facilitate benign self-perception and improve performance (Hirsch, Clark, Williams, Morrison, & Matthews, 2005; Hirsch, Meynen, & Clark, 2004). If Conway and Pleydell-Pearce (2000) are correct in their conceptualisation of the links between self and autobiographical memory, then self-imagery and autobiographical memory are an obvious target for investigation and provide the focus of the current study.

We designed the experiment reported in this paper to test whether holding a positive or negative image in mind during a speech would affect the retrieval of autobiographical memories to positive, negative and neutral word cues using the autobiographical memory task (AMT). The AMT is a cognitive-experimental paradigm that has been used extensively to investigate autobiographical memories in depression (e.g. Williams & Broadbent, 1986). In our study, high socially anxious participants gave two speeches, one while holding a positive image in mind and one while holding a negative image. They were tested on two separate occasions about 1 week apart.

There were two hypotheses. First, we predicted that memories retrieved to words that were congruent with the image would be retrieved faster than incongruent memories. This would mean that negative memories would be retrieved faster following the speech given while holding a negative image and the converse would be true for positive memories (faster latencies to retrieve positive memories following the speech holding a positive image). The second hypothesis functioned as a method check to substantiate any significant findings on the retrieval latency of autobiographical memories. Hirsch, Meynen, and Clark (2004) showed that when high socially anxious participants held a negative image in mind during a conversation, they felt more anxious, used more safety behaviours, rated their performance worse and underestimated their performance more than when they held a positive image in mind. We were using a different task (giving a speech) and a different method of eliciting the image (see the Procedure), and therefore the second hypothesis represented a partial replication of Hirsch et al.'s study. We predicted that when participants held a negative image during the speech they would feel more anxious, give worse ratings of performance and be rated worse by an independent assessor than when they gave a speech holding a positive image in mind.

Section snippets

Participants

Seventy undergraduate students from the University of Southampton were screened using the Fear of Negative Evaluation Scale (FNES: Watson & Friend, 1969). Twenty high socially anxious participants, who scored 17 or above (as recommended by Hirsch et al., 2004), took part in the study in exchange for course credits. Participants scoring 20 or above on the Beck Depression Inventory-II (BDI-II) were excluded from the study (n=4) because depressed individuals have shorter response latencies for

Success of the imagery manipulation

In order to evaluate the success of the imagery manipulation, we examined whether positive and negative images differed either in vividness, or in the amount of time the participants held the images in mind during the speeches. The mean vividness rating for negative images was 66 (SD=12.31) before the speech, and 58 (SD=19.08) during the speech. The mean vividness rating for positive images was 64 (SD=15.01) before the speech, and 51.5 (SD=16.31) during the speech. Image vividness was analysed

Discussion

The principal aim of this study was to look at the effect of self-imagery on the retrieval of autobiographical memories. A subsidiary aim was to see if we could replicate Hirsch et al.'s (2004) results using a different task and a different procedure. As the replication partly constituted a method check on the procedure, let us begin by looking at this part of the study first. Our results closely matched those obtained by Hirsch et al. When participants held a negative image in mind, they were

References (44)

  • R.M. Rapee et al.

    Mental representation of observable attributes in people with social phobia

    Journal of Behavior Therapy and Experimental Psychiatry

    (2006)
  • R.M. Rapee et al.

    The effects of videofeedback on the self-evaluation of performance in socially anxious subjects

    Behaviour Research and Therapy

    (1996)
  • R.M. Rapee et al.

    A cognitive-behavioural model of anxiety in social phobia

    Behaviour Research and Therapy

    (1997)
  • J. Spurr et al.

    Self-focused attention in social phobia and social anxiety

    Clinical Psychology Review

    (2002)
  • L. Stopa et al.

    Cognitive processes in social phobia

    Behaviour Research and Therapy

    (1993)
  • A.D. Baddeley et al.

    Working memory and the vividness of imagery

    Journal of Experimental Psychology

    (2000)
  • A.T. Beck et al.

    Beck depression inventory manual

    (1987)
  • G.D.A. Brown

    A frequency count of 190,000 words in the London–Lund corpus of English conversation

    Behavioural Research Methods Instrumentation and Computers

    (1984)
  • D. Clark et al.

    A cognitive model of social phobia

  • M.A. Conway et al.

    The construction of autobiographical memories in the self memory system

    Psychological Review

    (2000)
  • A. D’Argembeau et al.

    Phenomenal characteristics of autobiographical memories for social and non-social events in social phobia

    Memory

    (2006)
  • Hirsch, C. R., & Clark, D. M. (in press). Imagery special issue: Underestimation of auditory performance in social...
  • Cited by (61)

    • Improving the reporting of sport imagery interventions with TIDieR

      2022, Asian Journal of Sport and Exercise Psychology
      Citation Excerpt :

      That greater MG-M imagery ability may also act as a buffer against the effects of debilitative images (Quinton et al., 2019) is an important consideration when developing an intervention for this population. Student-athletes with high trait competitive anxiety may be particularly prone to experiencing involuntary images of a debilitating nature when faced with a stressful situation, which in turn, may have a quick and detrimental impact on their performance and well-being and lead them to using maladaptive emotional regulation strategies that can have long-term implications for their mental health (Nordin & Cumming, 2005; Chiupka et al., 2012; Moscovitch et al., 2011; Hirsch et al., 2003; Hirsch et al., 2004; Stopa & Jenkins, 2007). For this reason, PACING incorporates cognitive and behavioral emotion regulation strategies within the intervention as well as making use of evidence-based approaches such as written emotion disclosures to support athletes in generating their own scripts (Konig et al., 2014).

    • Avatar embodiment experiences to enhance mental health

      2020, Technology and Health: Promoting Attitude and Behavior Change
    View all citing articles on Scopus
    View full text