Andrew Menatti and Bethany Neczypor presented at a symposium at the Association for Behavioral and Cognitive Therapies annual convention Nov. 20-23 in Philadelphia.
Menatti and Neczypor are Psychology graduate students at Ohio University. Dr. Justin Weeks, Assistant Professor of Psychology, and Dr. Julie Suhr, Professor of Psychology, were co-authors. The symposium was on “You’re Making me Anxious”: Development, Validation, and Creative Implementation of Social Evaluation Stimuli.
Experimental utilization of social evaluation stimuli has been invaluable to the field’s continually emerging knowledge of social anxiety. The feared situations of socially anxious individuals involve a complex social world with innumerable facets, and experimental designs are essential to isolate and examine the myriad factors which can comprise social situations. Even relatively simple stimuli, when implemented appropriately, have been informative. For example, social evaluative words have been used in Stroop paradigms to demonstrate that socially anxious individuals selectively attend to cues that they are being perceived negatively (e.g., Gilboa-Schechtman, Foa, & Amir, 1999). This symposium discussed novel stimuli that have been designed and validated for use in social anxiety research. These stimuli are designed to be used in innovative ways, to enhance ecological validity of experimental designs, and to allow for more fine-grained analyses of contributing factors to social anxiety. In light of recent evidence that socially anxious individuals tend to fear positive, as well as negative evaluation (Kashdan, Weeks, & Savostyanova, 2011), several of the stimuli include positive as well as negative and neutral social cues. Stimuli presented in this symposium will also broaden the scope of social stimuli beyond the positive-negative continuum by manipulating other dimensions of social experience, such as social dominance and eye gaze.
The symposium began with a discussion of emotional video clips of an audience that vary on a continuum from critical to accepting. The advantages of such dynamic, externally valid stimuli for information-processing and eye-tracking research were considered. Next, the importance of social dominance in relation to evolutionary conceptualizations of social anxiety (e.g., Gilbert, 2001) was discussed, and research was presented on faces and statements that have been standardized with regard to social dominance. Applications of these stimuli also were reviewed. Next, the advantages of dynamic stimuli in approach-avoidance research was considered. Emotional video clips used in this research were presented, along with their creative implementation. The results of associated studies and their implications in the field of social anxiety also were discussed. In the final talk, emotional photographs of faces, modified for direction of gaze (i.e., direct versus averted gaze), were presented. The potential applications of these stimuli and the importance of standardized stimuli for information-processing research will be discussed. The symposium concluded with a discussion of the importance, and potential implications of standardized, ecologically valid stimuli in social anxiety research. The discussion was led by Dr. Eva Gilboa-Schechtman, a social anxiety expert with expertise in attentional processes, social rank, and non-verbal cues in relation to social anxiety, as well as experience in the use of social stimuli in experimental designs.
Validation of Structured Dynamic Clips of an Audience: Novel Experimental Stimuli for Social Anxiety Research
Theoretical models of social anxiety propose that preferential attention towards social threat serves to maintain beliefs that the social world is threatening (Heimberg, Rapee & Brozovich, 2010). Previous research evaluating social information processing relies primarily on photographic emotional stimuli. However, the social world is not comprised of static images. Social interactions require engagement with a dynamic compilation of positive, neutral, and negative stimuli. Thus, the current study validated a recently developed standardized set of nine emotional video stimuli (Skidmore Clips of Emotional and Neutral Expressive Scenarios; SCENES). The stimuli are 9 5-minute scenes of 10 people acting as members of an audience. For each scene every audience member received a “script,” such that they were instructed to appear either “critical,” “neutral,” or “accepting” of a presenter. Beginning at Scene 1, the scenes range from “unequivocally approving” (i.e., all actors were instructed to appear accepting) to Scene 7, “unequivocally disapproving” (i.e., all actors were instructed to appear critical). Scenes 2-6 were scripted to represent a continuum from low to high criticalness. Scenes 8 and 9 showed varying constitutions of a neutral audience.
The current study evaluated stimuli validity via an internet-based survey administered with Qualtrics, and participants were recruited via Amazon’s Mechanical Turk (MTurk). The sample included 383 (51.4% Caucasian; 58.9% female; M age = 34.84, SD age = 12.53). Each participant rated three randomly selected scenes on a nine-point Likert scale ranging from “extremely accepting” to “extremely critical.” Given the random selection of SCENES, each scene was rated by between 84 and 93 participants.
Study results provide compelling validity data for the SCENES videos. One-way ANOVA revealed a significant effect of scene F(8,777)=38.80, p<.001, η2=.285. Tukey’s post-hoc analyses revealed that participants rated Scene 1 as significantly more accepting than any other scene and Scene 7 as significantly more critical than any other scene (all ps<.05). As expected, ratings of the criticalness of the audience increased from scenes 1 to 7, Ratings of scene criticalness were significantly different from one another with few exceptions. In short, results provide support for the validity of these stimuli as depicting the expected range of criticalness. Such standardized dynamic emotional stimuli will potentially be an important step towards the execution of externally valid information-processing research.
“Where do They Rank”?: Evaluations of Perceived Social Dominance of Written and Facial Stimuli
According to psycho-evolutionary models, social anxiety facilitates interactions within dominance hierarchies (e.g., Trower, et al., 1990). Thus, social anxiety operates as a warning signal that others may be threatening, and triggers submissive responses which serve to appease dominant group members. Accordingly, social anxiety is closely related to perceptions of social dominance; however, there is a lack of experimental research in this area. In the current study, social stimuli were validated to facilitate research on the relations between social dominance and social anxiety. Specifically, this study investigated perceptions of social dominance of faces and statements.
Statements were developed by the first author and consisted of a total of 20 responses to 5 “get to know you” prompt questions. 12 responses were designed to be highly socially dominant, and 12 were designed to be less socially dominant (and yoked with regard to the high dominant statements in structure and content). Photographs were headshots obtained from the Center for Vital Longevity’s Face Database (Minear & Park, 2004). 16 initial faces were selected based on results from a preliminary survey; specifically, 4 were selected from each of the following 4 groups: Caucasian (1) males and (2) females, and African-American (3) males and (4) females. The decision to limit the facial stimuli to these racial backgrounds was based on results from a pilot study; fewer photographs were available for other ethnic groups, and these photographs lacked variability in social dominance.
A total of 159 participants (50.3% female; 50.9% Caucasian) were recruited via Amazon’s Mechanical Turk. Participants rated 6 statements (one for each prompt question) and 4 faces that matched their self-identified race/gender. Ratings of dominance, popularity, leadership, and competitiveness were averaged to create a “social dominance” composite score.
Social dominance differed as expected for responses to 4 of the 5 questions (all Fs >24.00; all ps<.001). Specifically, the statements designed to portray high levels of dominance were rated significantly higher on social dominance than were their yoked low-dominance statements (all ps <.001). Also, within each racial/gender group, 2 faces were rated as significantly more socially dominant than the other 2 faces (all Fs >5.80; all ps <.05). These stimuli may be utilized to create believable interaction partners in computer simulations, and may be effectively used to study the role of social dominance in social anxiety.
Dynamic Stimuli and Social Anxiety: Approaching Ecological Validity?
Research with static face stimuli has shown that social anxiety is related to increased avoidance impulses towards angry but also smiling faces (Lange, Keijsers, Becker, & Rinck, 2008). It is not clear, yet, what evaluation motivates especially the latter: Fear of positive evaluation or interpreting the smile as mockery. To investigate this, in a first study, 86 participants were shown 64 previously validated video clips of 8 different actors (4 female) who made 8 friendly, 8 insulting, 8 ambiguous or 8 neutral statements directed at the participants in a randomized order. After each clip the participants had to rate the impact it would have if the person had confronted the participant with his/her statement in real life on a paper form. They rated each clip with regard to their own emotional state on a 9-point Likert-scale ranging from very positive to very negative, and their arousal ranging from very calm to very aroused. The clip presentations were constructed in such a way that two specific actors (one female) only made negative statements, while two others said only positive, two only neutral or two actors both positive and negative things. Then, neutral photos of the same actors, were shown in an Approach-Avoidance Task (AAT). In the AAT participants used a joystick to pull the actors’ neutral looking faces towards themselves (approach) or push them away (avoidance) from themselves, depending on the background color. Results indicated that degree of social anxiety was positively correlated with avoidance tendencies in response to the photos of actors who had been insulting but not for actors who had been friendly. In a second study, the same clips and photos of the actors were used to assess avoidant body sway as well as freeze responses while participants were watching the stimuli while standing on a stabilometric force platform. The data collection of the latter study is ongoing and the results will be presented. The role of true negative but also of positive evaluation in social anxiety as well as the advantages of these kind of standardized dynamic stimuli will be discussed.
Manipulation of Emotion and Gaze Direction in a Set of Facial Stimuli: Initial Validation
Validation of facial stimuli is a critical step that is necessary in order to rule out plausible alternative explanations for significant results. Characteristics of facial stimuli should be empirically verified so that it can be reasonably assumed that participants will experience the faces in a predictable fashion. One important aspect that may be relevant to how participants experience the emotion communicated by a face is eye contact, however whether a person experiences a face as making eye-contact may depend on certain individual difference variables such as social anxiety (Gilbert, 2001). For example, past research has shown that eye-gaze is best conceptualized as a cone of direct gaze which refers to the angular space that a “looker” can be looking pointing their eyeballs toward a “receiver”, and be perceived by the “receiver” as making eye contact (Gamer & Hecht, 2007). In the present study, a large bank of pictures of faces was altered using photo-editing software in order to adjust of the apparent gaze direction of each face (see Ando, 2002). The faces were intended to show either a happy, angry, or neutral expression. Current data is from a preliminary sample (N = 21, 76.2% female) and includes ratings of which emotion a face is likely showing, ratings of how intensely each emotion is being expressed, ratings of whether they perceive the image as making eye contact with them, and finally, if anything appears unusual or unnatural about the images (in order to control for effects produced by the photo-editing process itself). Data from a full sample (expected N = 50) will be available by May, 2014.
Results show that participants discriminated between pictures making eye contact and those not making eye contact with 96.7% accuracy. In addition, “happy” was the most-endorsed emotion for 100% of the pictures intended to be happy, “angry” was the most-endorsed emotion for only 55% of the pictures intended to be angry, while “neutral” was the most-endorsed emotion for 75% of the pictures intended to be neutral. In pictures where there was congruence between the intended emotion and most-endorsed emotion, there was no difference in perceived intensity of the emotional expression between pictures making eye contact and pictures not making eye contact t(83) = -.99, p = .33. Chi-square analyses also showed that photo editing had no effect on participants’ perception of whether an image seemed unusual or unnatural χ2 (8) = 8.26, p = .41.
Additional findings from the full sample along with basic and applied applications of the stimuli will be discussed.
Comments