“Don’t worry, you are not alone”: Using Socially Interactive Agents to support Emotion Regulation
Non-effortful implicit emotional awareness (IEA, e.g. self-awareness/ consciousness) and effortful explicit emotional awareness (EEA, e.g. self-compassion) jointly influence emotion regulation (Guyrak et al., 2011).
Self-compassion (SC) reduces personal distress (Duarte et al., 2016), can be more effective than cognitive reappraisal (Diedrich et al., 2014), and mediates childhood maltreatment and later emotional dysregulation (Vetesse et al., 2011). SC develops through implicit co-regulation in childhood and may depend on social identity – interacts with cognitive and emotional appraisal in conflict situations (Higgins, 1996). How can IEA be supported in adults to influence dysfunctional emotional experience?
Study 1: Shaming socially interactive agents (SIA) during conflict situations. H1: Mediation social identity (moderated by self-consciousness)→ cognitive emotional regulation → emotional experiences (boredom) )(see model in Results).
Avatar-person similarity (Lee et al, 2015) and self-awareness (SE) mediate the effect of SIA on IEA, but identification may have a negative effect (Hooi et al., 2017). How does SIC degree of similarity influence self-awareness; (and SC)?
Study 2: We test the effect of similarity on implicit (II, affinity, rapport; Seymour et al.,2021) and explicit (EI, appearance; Downs et al., 2017) perceived identification. H2: There will be a linear effect on perceived similarity starting from 0% to 100% condition. H3: EI will be most affected by the 0%-condition, and II by 50%-condition.
Study 1: SIA shames participants (N = 57) during an interview situation to increase threat, shame and its regulation. Online, user shortly writes reasons for promotion. Shaming: 3 times turning down -“Maybe I didn’t explain right”. SIA-behaviour defined in VisualSceneMaker.
● Identity (Cheek et al., 1994)
● Self-consc./awareness (Scheier et al., 1985)
● Cog. appraisal of thread (Gaab, 2009)
● Boredom (Baratta & Spence, 2015)
Study 2: Three participant-avatars compared regarding similarity, identity and affinity. Adjusted 5 main features (using MetaHuman) for facial identification: chin shape, jaw width, eye shape, lip thickness, nose shape (Abudarham & Yovel, 2016).
● 0% max. similarity vs.
● 50% vs. 100%-condition