University of Central Florida Undergraduate Research Journal - Perception of Facial Expressions in Social Anxiety and Gaze Anxiety
US tab


Social anxiety disorder (SAD) is among the most common mental health issues in the United States. Those with SAD experience an intense fear of scrutiny during social interactions. These individuals may possess beliefs that others are negatively evaluating them, that they appear anxious and fearful to others, or that their behavior is outwardly offensive (American Psychiatric Association [APA], 2013). Social anxiety can lead to limited personal relationships and difficulties in the workplace, and it is highly correlated with depression and substance abuse (APA, 2013; Cooper, Hildebrandt, & Gerlach, 2014).

Engagement in avoidance behaviors is one method of mitigating social anxiety symptoms. These behaviors involve inflexibly avoiding any situation which provokes fear. Some examples include turning down a job that requires public speaking, avoiding speaking on the phone, or avoiding asking someone out on a date. The problem with avoidance is that it may contribute to the maintenance of anxiety symptoms due to operant conditioning processes (Feather, 1963). Not facing fearful situations reinforces avoidance, yet the anxiety about the situation is still present. Recent research has shown that greater levels of avoidance correlate with greater levels of anxiety and distress (Panayiotou, Karekla, & Mete, 2014).

Avoidance is not limited to social situations. It can also be demonstrated by the avoidance of specific stimuli. Mutual gaze (direct eye contact) is a commonly reported fear in persons with SAD, and research has shown that individuals with high social anxiety tend to fixate less on the eye region compared to healthy controls (Schulze, Renneberg, & Lobmaier, 2013). Although avoidance of gaze is a behavior demonstrated across multiple clinical disorders, how it influences other behaviors is uncertain. The current study aims to explore the relationship between social anxiety, gaze avoidance, and the perception of onlooker emotion. More specifically, is the avoidance of eye contact by individuals who are socially anxious related to how the facial expression of an onlooker is perceived?

Gaze Behavior in Social Anxiety

Where we direct our eyes is an indicator of where our visual attention is focused. During social interactions, our eye movement is used as a conversational cue; for example, eye movement provides social facilitation and information on whose turn it is to speak (McCarthy, Lee, Itakura, & Muir, 2006). Eye contact also communicates information about dominance, level of interest in the speaker, attention, comfort level, and emotional state (Schulze et al., 2013; Senju & Johnson, 2009; Weiser, Pauli, Alpers, & Muhlberger, 2009). Senju and Johnson (2016) found that engaging in eye contact—or even viewing pictures of faces with the eyes directed at the observer—may activate a number of brain structures associated with social information processing, such as the fusiform gyrus and the amygdala. They found that the activation of this "social brain" may increase performance on socially driven tasks, such as gender discrimination (Senju & Johnson, 2009). This suggests that the initiation of eye contact serves as an important priming stimulus to prepare observers for the processing of social information.

Considering that gaze is a method of exchanging social information, it is not surprising that clinical disorders involving social impairment involve gaze impairment as well. Maladapted gaze behaviors can include avoiding eye contact, fixating on irrelevant stimuli in the visual field, or inefficiently scanning for details. This list is not exhaustive and different forms of these manybehaviors manifest themselves in SAD, autism, schizophrenia, and Williams syndrome (Guillon, Hadjikhani, Baduel, & Roge, 2014; Schulze et al., 2013; Vaidyanathan et al., 2014). Eye tracking studies have shown that individuals with high levels of social anxiety traits will fixate their visual attention less on the eye region compared to the rest of the face and that this avoidance is preceded by hypervigilant scanning for negative information (Schulze et al., 2013; Terburg, Arts, & Honk, 2012). Hypervigilance is an increased sensitivity in detecting specific negative stimuli. Participants in these studies reportedly scan for negative social information more intensely and then are quick to avoid it when it is perceived (Terburg et al., 2012).

Despite a recent increase in the use of eye tracking to assess attention bias, research regarding gaze avoidant behavior by individuals with SAD is inconclusive (Schulze et al., 2013). Whether eye contact is actually avoided may be moderated by other factors in addition to anxiety, such as gender of the speaker, cultural context, or distance between speakers (Weiser, Pauli, Grosseibl, Molzow, & Muhlberger, 2010). If the avoidance of eye contact by persons with SAD is dependent on these moderating factors, the presence of these moderating factors may explain why results concerning gaze avoidance have not consistently shown avoidance of the eye region across all social situations.

Even when avoidance is not present, increased physiological arousal has been noted in response to eye contact. Weiser and colleagues (2009) found that socially anxious individuals had higher cardiac responses to eyes directed towards them when compared to individuals who were less socially anxious. In a similar study, Schneier, Pomplun, Sy, and Hirsch (2011) compared the neural response to viewing pictures of faces with the eyes either directed at the observer or with the eyes averted away. Participants who met criteria for SAD were compared to a non-anxious control group. Consistently, direct gaze correlated with greater activation of areas in the brain associated with self-monitoring, selfreferential processing, and the processing of emotional information. After treatment, these activation responses decreased proportionally in intensity. Self-monitoring has particularly interesting clinical implications because individuals with SAD exhibit heightened awareness of their outward appearance (APA, 2013).

Adverse reactions to eye contact are not the only aspects in which gaze is influenced by social anxiety. Individuals experiencing social anxiety might be responding to direct eye contact even when direct eye contact is not occurring. Honma (2013) demonstrated this effect through selfreports and physiological responses. Participants in the experiment were placed face to face and designated to play the role of either the viewer or the perceiver. The viewer was instructed to look at the perceiver's eyes in one condition or at a set point in between the eyes in a separate condition. The perceiver was then asked to report when they perceived eye contact and had their pupil diameter measured during the task. This design allowed experimenters to measure the spatial field in which gaze is perceived as eye contact. Higher selfreported social anxiety on the Liebowitz Social Anxiety Scale (LSAS) was correlated with greater spatial range of eye contact perception. Additionally, perception of mutual gaze was correlated with greater amounts of pupil dilation even when eye contact was not made. Not only did socially anxious participants perceive greater amounts of eye contact than what actually occurred, but these individuals demonstrated heightened physiological arousal in response to this perceived eye contact.

Another study comparing individuals diagnosed with SAD to healthy controls found that increased spatial response to eye contact perception was only present when more than one viewer was looking at the perceiver (Gamer, Hecht, Seipp, & Hiller, 2011). Possessing a greater spatial range in which eye contact is perceived could contribute to the feeling of being observed. This feeling could help explain why individuals with SAD typically report fears of being observed during mundane behavior, such as eating (APA, 2013).

Facial Recognition in Social Anxiety

While previous literature supports both the misperception and avoidance of gaze by persons with SAD, it remains unclear what function gaze avoidance serves in the maintenance of anxiety. As with most avoidance behaviors, it is unlikely that this behavior helps to relieve anxiety symptoms in the long term. One possibility is that the fear of mutual gaze leads to the misinterpretation of important social information communicated nonverbally through the face or eyes. The literature supports this notion by showing that the perception of facial expressions is biased towards negative interpretations in individuals with SAD (GutierrezGarcia & Calvo, 2014; Mohlman, Carmin, & Price, 2007; Yoon & Zinbarg, 2007).

Another possible explanation for the tendency of persons with SAD to interpret social information negatively is that individuals experiencing high social anxiety operate with a lower threshold for detecting negative stimuli. That is to say, socially anxious individuals are more likely to detect negative stimuli at lower levels of intensity that might otherwise be ignored by individuals who are not socially anxious. Frenkel and Bar-Haim (2011) found support for this hypothesis when evaluating neural responses to small changes in the level of fear intensity in faces. Pictures of faces that gradually expressed a more fearful facial structure were presented while participants' brain activity, in the form of event related potentials (ERPs), were measured. ERPs provide a measurable response in the brain that are highly sensitive to the specific events of interest (e.g., gradual changes in facial expressions). Unlike the control group which had a graded response, those who were socially anxious did not demonstrate much discrimination in brain response to subtle changes in facial structure. Low levels of fear were processed with similar intensity as high levels of fear. The socially anxious group also had a lower threshold for identifying the fearful facial expressions than the control group. Similar results have been found in regards to anger. Socially anxious participants detected angry expressions at lower intensity levels than both a control group and a group of individuals with depression (Joormann & Gotlib, 2006).

Detecting emotional states at lower levels of intensity may be related to previous findings that show socially anxious individuals have a tendency to judge ambiguous information as negative (Yoon & Zinbarg, 2007). For instance, although socially anxious individuals can identify angry facial expressions with greater accuracy than controls, they are more likely to misidentify neutral faces as angry (Mohlman et al., 2007). If a lower threshold for threat detection is operating, it makes sense that ambiguous information (such as neutrality) triggers the misidentification of anger or fear during such recognition tasks. Regardless, an attentional bias that focuses on negative information is likely operating when individuals high in social anxiety process information.

Considering that social anxiety appears to influence face perception, it is important to account for how facial expressions are processed. The region of the face that is the focus of attention might depend on both the emotional state of the face being viewed and that of the viewer. A greater proportion of visual attention is spent on the eyes when the face is perceived as negative (Scheller, Büchel, & Gamer, 2012) and the experience of embarrassment or social anxiety in the viewer is correlated to even greater attentional biases involving the eye region (Darby & Harris, 2010; Gutierrez-Garcia & Calvo, 2014). Hills and Lewis (2011) have shown that mood is also related to preferential processing of the eye region. Participants induced to feel sad failed at detecting subtle changes in the eye region, while those induced to feel happy performed better at detecting eye changes than detecting changes on the rest of the face.

It seems likely that processing of the eye region is of preferential importance when evaluating the overall expression of a face with eyes, perhaps functioning as the most salient emotional cue. However, researchers typically do not examine the perception of the eye region independently–most likely due to the loss of contextual information contributed by the rest of face. Even though the literature suggests socially anxious individuals exhibit hypervigilance and avoidance towards the eye region (Terburg et al., 2012), few studies have explored how gaze avoidance and eye region perception interact when processed by socially anxious individuals (Horley, Williams, Gonsalvez, & Gordon, 2004).

The current study will evaluate the perception of emotion by participants high in self-reported social anxiety and gaze avoidant behavior using a facial recognition task. Performance on this task will be measured as a participant's accuracy at labeling emotions and identifying how intense they perceive those emotional expressions to be. The study will also explore the differences in perceiving emotion within only the eye region when compared to perceiving emotions in the entire face.

Materials and Methods >>