Social anxiety and interpretation biases for facial displays of emotion: Emotion detection and ratings of social cost

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

Abstract

The current study assessed the processing of facial displays of emotion (Happy, Disgust, and Neutral) of varying emotional intensities in participants with high vs. low social anxiety. Use of facial expressions of varying intensities allowed for strong external validity and a fine-grained analysis of interpretation biases. Sensitivity to perceiving negative evaluation in faces (i.e., emotion detection) was assessed at both long (unlimited) and brief (60 ms) stimulus durations. In addition, ratings of perceived social cost were made indicating what participants judged it would be like to have a social interaction with a person exhibiting the stimulus emotion. Results suggest that high social anxiety participants did not demonstrate biases in their sensitivity to perceiving negative evaluation (i.e. disgust) in facial expressions. However, high social anxiety participants did estimate the perceived cost of interacting with someone showing disgust to be significantly greater than low social anxiety participants, regardless of the intensity of the disgust expression. These results are consistent with a specific type of interpretation bias in which participants with social anxiety have elevated ratings of the social cost of interacting with individuals displaying negative evaluation.

Introduction

Theoretical models of social phobia suggest that biased information processing contributes to the etiology and maintenance of social anxiety (Clark & Wells, 1995; Rapee & Heimberg, 1997). These models posit that individuals with elevated social anxiety tend to demonstrate negative biases in processing social cues that are indicative of negative evaluation. Data have largely supported theoretical claims that these individuals demonstrate perceptual and processing biases, such as biased attention to social threat (Mogg & Bradley, 2002; Mogg, Philippot, & Bradley, 2004), excessive attention to internal physiological cues (Bogels & Mansell, 2004), and negative interpretations of ambiguous social events (Amir, Foa, & Coles, 1998; Foa, Franklin, Perry, & Herbert, 1996). Additional research supports the importance of these biases in maintaining symptoms, given that interpretation biases have been shown to be reduced following effective treatment (Foa et al., 1996; McManus, Clark, & Hackmann, 2000; Wilson & Rapee, 2005). Thus, further clarifying the nature of these biases will contribute to better understanding the etiology and maintenance of social anxiety, and may further inform treatment.

Traditionally, the investigation of interpretation biases has focused primarily on the assessment of potential threat (e.g., whether the stimulus/scenario is negative, likelihood of a negative outcome). Studies of this type of interpretation bias have repeatedly revealed that individuals with social anxiety interpret ambiguous social scenarios more negatively than do controls (Amir et al., 1998; Constans, Penn, Ihen, & Hope, 1999; Stopa & Clark, 2000). For example, individuals who are socially anxious may be more likely to interpret the scenario “you walk into the break-room at work and the conversation stops” as reflecting something negative about them in contrast to a neutral or positive interpretation (e.g., “the conversation had reached its natural end” or “they were looking forward to hearing what you had to say”). Similarly, negative interpretation biases have been demonstrated for mildly negative (Lucock & Salkovskis, 1988; Vassilopoulos, 2006), as well as positive (Vassilopoulos, 2006), social events. Clearly, given the ambiguous nature of the majority of social cues one encounters in daily life (e.g., a friend's distracted glance could indicate either that they are either preoccupied [non-threat] or disinterested [threat]), biases to negatively misinterpret social cues can serve to maintain anxiety related to interacting with the social world.

In addition to the estimate of perceived threat of social cues, the interpretation process also involves estimates of the extent of emotional cost associated with these social cues (i.e., the perceived impact of the social event on the individual). Existing data suggest that individuals with social phobia estimate the cost of ambiguous (Foa et al., 1996) and negative (McManus et al., 2000; Vassilopoulos, 2006) social events to be greater (i.e., more negative) than non-anxious individuals. Thus, anxiety may be further maintained by the interpretation that even mild (or unclear) social disapproval is ruinous. Research examining mediators of treatment change in social phobia supports the utility of distinguishing between estimates of probability of negative outcomes and the cost of negative outcomes (e.g., Foa et al., 1996; McManus et al., 2000), although which indicator is more important for change still remains unclear. Therefore, to fully understand the role of interpretation biases in social anxiety it is important to assess both the interpretation of the valence of the social stimuli and the emotional cost of such stimuli for that individual.

Although early research on interpretation in social anxiety focused primarily on the interpretation of social scenarios presented as text (e.g., Amir et al., 1998; Constans et al., 1999; Foa et al., 1996; Stopa & Clark, 2000), more recent work has extended this line of research to studying the interpretation of emotional facial expressions (e.g., Mullins & Duke, 2004; Philippot & Douilliez, 2005; Richards et al., 2002; Winton, Clark, & Edelmann, 1995). Understanding the interpretation of facial expressions is important for understanding social anxiety given that facial expressions convey essential social information (Darwin, 1872; Ekman, 2003). Further, accurate decoding of emotional expressions is related to relationship well being. For example, Carton, Kessler, and Pape (1999) demonstrated that errors in decoding emotional expressions in faces were significantly correlated with low-relationship well-being and symptoms of depression.

To date, the literature on processing of facial expressions in social anxiety is sparse. On a basic level, Mullins and Duke (2004) found that social anxiety was not associated with errors in decoding facial expressions (labeling faces as happy, angry, fearful or sad), regardless of state anxiety. Further, these results were demonstrated for both intense and subtle displays of emotion. Using facial stimuli displaying intense emotional expressions, Winton et al. (1995) did not find evidence for differences between individuals that were high and low in social anxiety in their ability to discriminate negative vs. neutral facial expressions (label angry, sad, disgust, contempt, fear and neutral faces as either “negative” or “neutral”). However, their results showed that individuals with elevated social anxiety exhibited a response bias to report that all facial expressions were “negative.” In another study conducted by Richards et al. (2002), participants viewed facial expressions representing a mixture of two emotions (e.g., 10% anger and 90% disgust; 30% fear and 70% sadness, etc.). Participants were then asked to label the emotional expression. The authors concluded that when fear was one of the two component emotions in the face, participants with elevated social anxiety classified more of the faces along the continuum as expressing fear relative to low-anxious controls. However, there was not a significant interaction of anxiety level with the percentage of fear displayed in the face. These results show that when presented with faces representing a mixture of fear and surprise or fear and sadness, individuals with elevated social anxiety are more likely to label the faces as fear. The authors argue that the results were not the result of an overall response bias to label emotional expressions as fearful as group differences were not found for the likelihood of labeling faces representing mixtures of disgust/anger, anger/happiness, happiness/surprise, or sadness/disgust as representing fear. Importantly, although this study addresses the recognition of emotional expressions, the stimuli employed in the study were blends of two different emotions (e.g., surprise/fear, anger/happiness) rather than nuances of pure emotional expressions. Thus, this study fails to address the impact of subtle emotional cues, such as low-intensity negative expressions. Utilizing such ambiguous indicators of negative evaluation (e.g., low-intensity expressions of threat) is particularly relevant to understanding interpretation biases in social anxiety. Importantly, the impact of such subtle emotional cues was investigated by Philippot and Douilliez (2005). This study employed morphed facial expressions of varying intensities, morphing between neutral expressions to emotional expressions, thereby assessing the impact of subtle emotional cues (0%, 30%, 70%, and 100% emotion). Results did not show group differences among individuals with social phobia, individuals with other anxiety disorders, and non-anxious controls on decoding accuracy (which emotion received the highest intensity rating on continuous emotion scales). Philippot and Douilliez (2005) speculated that despite a lack of an explicit evaluative bias, these biases may manifest in implicit or automatic processing as well as ratings of the potential importance (or meaning) of the stimuli to oneself (i.e., perceived emotional cost).

Therefore, the primary goal of the current study was to further the understanding of the interpretation of facial expressions conveying negative evaluation in social anxiety. Building from previous data and existing theory, three tasks were utilized to: examine perceptions of negative evaluation in facial expressions (tasks 1 and 2) and perceived cost (task 3) of these facial expressions, and examine the perception of negative evaluation at both long (unlimited presentation time; task 1) and brief (60 ms, task 2), stimulus durations. Given the aim of addressing interpretation of emotional facial expressions, and the fact that intense facial expressions leave less room for interpretation, stimuli were created for the current study representing a range of emotional intensities. Disgust and happy facial expressions were morphed with neutral expressions to create stimuli of varying intensities (e.g., 20%, 30%, 40% disgust, etc.). Disgust expressions were used in the current study for numerous reasons. First, as noted by Amir et al. (2005) disgust expressions are particularly relevant to social anxiety as they convey rejection. Indeed, disgust expressions provide a strong match to cognitive–behavioral models regarding factors that elicit social anxiety. That is, disgust expressions can be seen as “external indicators of negative evaluation” which are a core component of the theoretical model by Rapee and Heimberg (1997). Finally, results of a recent imaging study documented that social phobia participants displayed differential neural activation when processing disgust facial expressions in comparison to matched controls (Amir et al., 2005).1 Happy expressions were also included to determine whether any interpretation bias detected would be specific to disgust or would be evident for a positive emotion as well.

Based on theoretical models of social anxiety (Clark & Wells, 1995; Rapee & Heimberg, 1997) and the existing literature on interpretation biases that has used text (Amir et al., 1998; Constans et al., 1999; Foa et al., 1996; Stopa & Clark, 2000), we hypothesized that individuals with elevated social anxiety would show a bias to label faces as portraying disgust at a lower intensity than individuals with low social anxiety (i.e., show a heightened sensitivity to threat). This said, recent findings (Philippot & Douilliez, 2005) suggest that group differences may only emerge when participants are tested using brief stimulus durations (task 2) rather than at long stimulus durations (task 1). Finally, we hypothesized that individuals with elevated social anxiety would rate disgust faces (regardless of intensity) as more costly to interact with than would individuals with low social anxiety. As suggested above, happy faces were used as an emotion control; no specific predictions were made regarding potential group differences for the happy faces.

Section snippets

Participants

Participants for this study were 100 undergraduates who received research credit in introductory psychology courses for participation. Potential participants were identified via the psychology department's mass testing sessions. High social anxiety and low social anxiety groups were identified from this screening sample (n=697) based on quartile scores on the Brief Fear of Negative Evaluation (BFNE) scale (Leary, 1983). Individuals who scored in the top 25% (BFNE total ⩾40) were recruited as

Results

Means and standard deviations for the demographic variables are presented in Table 1. There were no significant differences between the high and low social anxiety groups in terms of age [t(98)=1.66, p=0.10, reffect size=0.17] or ethnicity (Caucasian vs. non-Caucasian) [χ2 (1, N=100)=0.17, p=0.68, reffect size=0.04].5

Discussion

Findings from tasks one and two failed to provide support for differences between individuals with high and low social anxiety in regards to their perceptions of negative evaluation in facial expressions (i.e., in faces portraying disgust). Specifically, the two groups were not found to differ significantly in their likelihood to label disgust expressions of varying intensities as “disgust”. Further, the same pattern of results was found for both unlimited and brief (60 ms) stimulus presentation

References (39)

  • L. Stopa et al.

    Social phobia and interpretation of social events

    Behaviour Research and Therapy

    (2000)
  • J.K. Wilson et al.

    The interpretation of negative social events in social phobia: Changes during treatment and relationship to outcome

    Behaviour Research and Therapy

    (2005)
  • E.C. Winton et al.

    Social anxiety, fear of negative evaluation and the detection of negative emotion in others

    Behaviour Research and Therapy

    (1995)
  • A.W. Young et al.

    Facial expression megamix: Tests of dimensional and category accounts of emotion recognition

    Cognition

    (1997)
  • N. Amir et al.

    Negative interpretation bias in social phobia

    Behaviour Research and Therapy

    (1998)
  • M. Biehl et al.

    Matsumoto and Ekman's Japanese and Caucasioan facial expressions of emotion (JACFEE): Reliability data and cross-national differences

    Journal of Nonverbal Behavior

    (1997)
  • E.J. Brown et al.

    Validation of the social interaction anxiety scale and the social phobia scale across the anxiety disorders

    Psychological Assessment

    (1997)
  • J.S. Carton et al.

    Nonverbal decoding skills and relationship well-being in adults

    Journal of Nonverbal Behavior

    (1999)
  • D.M. Clark et al.

    A cognitive model of social phobia

  • Cited by (70)

    • Interpretation bias in social anxiety: A systematic review and meta-analysis

      2020, Journal of Affective Disorders
      Citation Excerpt :

      That is, the degree of threat embedded in facial photographs may not be strong enough for socially anxious individuals to perceive them as threatening compared to other types of stimuli such as videos, sentences or scenarios. In addition, it has been suggested that facial expressions without context may be associated with confusion because not only anger but also disgust, contempt, and even happiness can be interpreted as signals of social rejection and hostility (Campbell et al., 2009; Jusyte and Schonenberg, 2014; Schofield et al., 2007). This suggests that using different emotional expressions without context to access threat interpretation may lead to discrepant results.

    View all citing articles on Scopus
    View full text