by Michael D. Anestis, M.S.
I came across a study this morning with startlingly impressive findings. What's more, the article was published in the most prestigious clinical psychology journal (Journal of Abnormal Psychology) and was written by several of my current colleagues at Florida State University and an FSU alum, and yet I did not know about it until just hours ago. This goes to show just how difficult it is to keep up with all of the research transpiring in this field, even when it takes place in the same building as my own office.
In this particular study, Brad Schmidt, Tony Richey, Julie Buckner, and Kiara Timpano (2009) investigated the potential utility of a computer-based intervention for social anxiety disorder (SAD) that aims to adjust the degree to which individuals with SAD attend to threat cues (e.g., facial expressions that might indicate others are evaluating them negatively). The rationale for this particular study was based on a rich, although not universally consistent, literature indicating that individuals with SAD demonstrate an attentional bias towards threat (e.g., Amir, Freshman, & Foa, 2002; Rapee & Heimberg, 1997). What this means is that individuals with SAD are hypervigilant towards threat, utilizing many of their attentional resources to scan their environment for evidence that they are currently being evaluated poorly. Individuals with high levels of social anxiety have been shown to detect greater amounts of negative audience behavior relative to individuals with low levels of social anxiety (Valjaca & Rapee, 1998).
An argument could be made that attending to threat cues is an adaptive tendency. After all, shouldn't we be aware of moments in which we are in danger? If the question is phrased in this manner, the answer would be yes; however, this question misses the point. Individuals with SAD are not attending to threats that put them in physical harm. Instead, because this disorder is characterized by a fear of negative evaluation, the focus is on cues that, while likely to cause discomfort, are not truly representative of actual danger. Additionally, constant hypervigilance towards threat serves to increase and sustain anxiety (Beck, Emery, & Greenberg, 1985). So the actual result of this attentional bias is an overestimation of the likelihood and severity of threat as well as increased and prolonged negative mood.
Schmidt and colleagues (2009), thus, set out to address the attentional biases of individuals with SAD by training them to control their attention and shift focus away from threat cues. In this study, the authors places 18 individuals with SAD into a treatment condition and 18 individuals with SAD into a control (placebo) condition. Both the treatment condition and the control condition consisted of a computer task that presented participants with a series of pairs of pictures, in addition to probes that prompted them to identify letters. More specifically, participants were told to look at a screen while a series of photos were presented to them. Prior to the presentation of each pair of photos, a fixation cross was displayed on the screen for 500ms. Two photos were then presented on the screen, one immediately above the placement of the cross and one immediately below it. Photos were either threat-based (a disgusted facial expression) or neutral (neutral facial expression). Additionally, in both conditions, one of the photos was followed by a probe (either a letter E or a letter F). When a probe appeared, the participant was told to indicate which letter had been presented. Participants were instructed to perform this task as quickly as possible without sacrificing accuracy.
The difference between the two conditions was based on the placement of the probe. In the treatment condition - individuals with SAD receiving attentional control training - the probe appeared in the place previously occupied by a neutral face in 80% of the pairings. In this way, even though participants in this condition were not told to look away from disgusted faces, they learned that, when they saw disgusted faces, they should look elsewhere to find the probe. As a result, seeing a disgusted face became a cue to look elsewhere. In the control condition, the probe appeared in the location previously occupied by the disgusted face as often as it appeared in the location previously occupied by the neutral face.
Now, at first glance, the therapeutic value of this type of task may seem minimal. How can looking at photos and identifying letters impact social anxiety? Well, this is where data comes in handy, as it helps us get past strong but inaccurate gut reactions. In this study, the authors measured SAD symptoms through self-report but also, more importantly, through a structured diagnostic interview, the most valid and reliable manner for reaching diagnostic decisions. Symptoms were measured pretreatment, post-treatment, and at a 4-month follow-up. Amazingly, the authors found that, at termination, 72% of individuals in the treatment conditions no longer met criteria for SAD compared to 11% of individuals in the control condition. Additionally, these benefits were maintained at 4-month follow-up. Even more astoundingly, only one participant in the treatment condition correctly guessed that he or she was receiving treatment. In other words, every person in the study except one believed that they were in the no-treatment condition, so the belief that they were being treated can not account for these effects.
Perhaps even more impressive than the magnitude of these findings was the amount of time it took to attain them. Treatment consisted of eight 15-minute sessions, typically twice per week. The average amount of time between pre- and post-treatment assessments was 36.8 days. Additionally, these effects were attained with almost no contact with an actual therapist. What this means is that the authors were able to help a substantial number of individuals with SAD remit from their illness in very little time entirely through a computer task that the individuals themselves believed was not even a treatment.
There are many benefits to research like this. First, even though there are already empirically supported treatments for SAD (see Lora Rose Hunter's PBB article on cognitive behavioral group therapy for SAD), not every individual responds to the front line treatment or has access to clinicians trained to administer empirically supported treatments. Secondly, creating effective, empirically supported online treatments might increase the degree to which individuals with SAD seek and attain proper treatment.
Ultimately, this study accomplished several admirable goals. First, it presented initial empirical evidence in support of a brief, inexpensive treatment for SAD that maintains treatment benefits well beyond the end of treatment. Second, it provided an example of a thorough assessment approach utilizing scientifically tested tools (see our Assessment Tools page for more examples). Finally, this study is a prime example of the need to take a scientific approach to clinical psychology. The most effective means for reducing symptoms of mental illness are not always the most philosophically appealing or intuitively obvious. The reduction of symptoms for clients is more important than intellectual satisfaction for clinicians, and science and data are the most effective means for ensuring that such reductions remain our highest priority.
To learn more about treatments for social anxiety disorder or methods for developing attentional control, please refer to our online store for recommended products.
Mike Anestis is a doctoral candidate in the clinical psychology department at Florida State University
Subscribe in a reader
