Watch what I do, not what I say I do: Computer-based avatars to assess behavioral inhibition, a vulnerability factor for anxiety disorders

Abstract

Behavioral inhibition (BI), a tendency to withdraw from or avoid novel social and non-social situations, is a personality trait which can confer risk for anxiety disorders. Like many personality traits, BI is often assessed via self-report questionnaires where respondents rate themselves for frequency of certain behaviors or feelings. However, questionnaires have inherent limitations, particularly in psychiatric populations where there may be unawareness of deficit. A viable alternative may be virtual environments, in which the participant guides an on-screen “avatar” through a series of onscreen events meant to simulate real-world situations. Here, we report on initial development of such an assessment tool, involving several onscreen scenarios with choice points where the participant can select from response options corresponding to inhibited or uninhibited behaviors. In two experiments involving over 300 college students, scores on the computer-based task were strongly correlated with BI scores attained through self-report questionnaire (r>.780, p<.001); this relationship held regardless of participant gender and experience with computer games. The results suggest that virtual environments may hold promise as alternative formats for assessment of personality traits in populations unsuited to traditional paper-and-pencil questionnaire formats due to psychopathology, limited attention span, or poor vocabulary and/or literacy skills.

Beyond symptom self-report: use of a computer “avatar” to assess post-traumatic stress disorder (PTSD) symptoms

“…The study also replicated prior findings (Myers et al, 2016) that task performance predicts participants’ self-reported tendency to engage in avoidant behavior, assessed by AMBI. The prior study examined a sample that was younger, more educated, and highly computer literate.…”

“…Although our prior article (Myers et al, 2016) found a strong correlation between task and AMBI scores (r=.783, r 2 =.613), we did not have preliminary data regarding correlations with PCL scores. Accordingly, we based power calculations on the sample size needed to detect a medium effect size (r=.3, r 2 =.09) at 80% power, if such an effect exists.…”

“…The computer-based task was as previously described (Myers et al, 2016); an open-access version of the task software is available at Open Science Framework www.osf.io/zf3jv. The task took about 10 minutes for participants to complete.…”

“…Following completion of the computer-based task, participants were given a short post-test questionnaire (Myers et al, 2016) which asked them to rate, on a four-point scale (none, a little, moderate, high), their familiarity with computers, whether they every played computer games (yes/no), and if so what type. Participants were also asked to rate how much they cared about what happened to the avatar in the game (not at all, a little bit, somewhat, a lot), and whether they made decisions in the game that were similar to the way they normally act in real life (not at all, a little bit, mostly, exactly like me).…”

“…Recently, we developed a computer-based platform in which the user guides an on-screen avatar through a series of scenarios representing novel social situations with defined choice points that might invoke avoidant behavior (Myers et al, 2016). In a sample of healthy young adults, task performance was strongly correlated with temperamental tendency for avoidant behavior, assessed via the Adult Measure of Behavioural Inhibition (AMBI; Gladstone & Parker, 2005), a scale which correlates with anxiety proneness (Gladstone & Parker, 2005) and with PTSD symptom severity in veterans (Myers, VanMeenen & Servatius, 2012).…”