Poster C34, Friday, August 17, 10:30 am – 12:15 pm, Room 2000AB

Individual differences in sarcasm perception: behavioral and eye tracking evidence

Kathrin Rothermich1, Mathew Fammartino2, Hana Kim1, Gitte Joergensen2;1East Carolina University, Greenville, USA, 2University of Connecticut, Storrs, USA

Social communication is usually complex and hardly ever straightforward. Often listeners must infer what speakers mean since what is said and what is meant can differ. Crucially, interpreting nonliteral language such as sarcasm or teasing involves complex cognitive skills such as the inference of mental states, the integration of paralinguistic cues, perspective taking, and empathy. Individual differences also play a role in a person’s ability to understand and interpret sarcasm, its appropriateness, and whether a particular use of sarcasm was perceived as amicable or hostile. For the current study, we invited young adults (N=15; mean age = 18.6 years, SD = 0.73) to watch brief videos of dyadic social interactions in which responders’ intentions were sincere or sarcastic. The videos were taken from a novel database (RISC; Rothermich & Pell, 2015) and include a multitude of cues such as prosody/intonation, body language, and facial expressions while keeping lexical information constant. Participants were also presented with several questionnaires assessing their social communication preferences (e.g., Sarcasm Self-Report Scale; Ivanko, 2004) and social anxiety (e.g., STATE-TRAIT Anxiety Inventory; Spielberger, 2010). During the experiment, participants identified the sincerity of the responder (sincere/sarcastic) and rated their likeability and friendliness on a 5-point Likert scale. At the same time, eye movements were recorded using an Eyelink 1000Plus eye-tracking device. The results demonstrate that overall accuracy in identifying sincerity is well above chance (91.96 % correct, SD = 2.68). The Likert scale results show that actors using sincere responses were significantly rated as more friendly and likeable (likeability: M= 4.33, SD = 0.87; friendliness: M= 4.46, SD = 0.70), compared to when using sarcastic responses (likeability: M= 1.83, SD = 0.84; friendliness: M= 2.00, SD = 0.91). Because sincere and sarcastic interactions contain the exact same lexical content, these results reveal the influence of nonverbal signals and prosody on speaker impression. Additionally, a significant correlation was found between the Sarcasm Self-Report Scale and friendliness ratings for sarcastic responses (R = .56, p < .05), indicating that participants who report using sarcasm regularly rate these responses as more friendly than those participants who rarely use sarcasm. Finally, the eye-tracking results show that participants spent significantly more time fixating faces when processing sarcastic interactions (M = 698, SD = 12.72) compared to sincere interactions (M = 629, SD = 17.05). Interestingly, fixation duration was negatively correlated with measured TRAIT anxiety (R = -.55, p < .05), indicating that participants with higher levels of social anxiety spent less time viewing sarcastic scenes, as predicted by the hypervigilance-avoidance hypothesis (Mogg et al. 1997). Taken together, the present data speak for an influence of social communication preferences and personality factors on the processing of sarcasm.

Topic Area: Meaning: Prosody, Social and Emotional Processes