Epistemic modeling of moral emotions and appraisals following third-party intervention in peer victimization

• In Event: 1-146 - Employing Quantitative Ethnography and Epistemic Network Analysis to Model Emotions and Discourse in Moral Development

Integrative Statement

When youth witness aggression towards peers, their responses—to console the victim, intervene aggressively, attempt a peaceful resolution, or ignore the situation—may stop the aggression, buffer its negative effects, or conversely, escalate and compound the harm to victims. Such actions stimulate self-reflection among adolescents (Opotow, 2004), which may influence future behavior through a rich network of moral emotions, social norms, and self-appraisal (Krettenauer, Colasante, Buchmann, & Malti, 2014; Malti, & Krettenauer, 2013; Smetana, Jambon & Ball, 2014). We interviewed adolescents to see how they felt and thought about their reactions after a peer had been targeted for aggression, harassment or discrimination. To demonstrate how Epistemic Network Analysis (ENA, Schaffer, 2017) can be applied to quantitative data, we focus primarily on emotion and appraisal ratings.
We asked 270 youth between 13 and 18 (26.0% African-American, 24.2% European-American, 24.5% Mexican-American, 25.3% Native-American) to describe times when they had attempted to (1) calm a victim’s emotions, (2) amplify anger, (3) avenge a peer’s victimization, and (4) forge a peaceful resolution. Interviewees identified the strength of emotions they experienced and explained why they experienced each one. They appraised their actions with respect to three criteria: (1) helpfulness, (2) how much they felt like a good friend, and (3) the percentage of peers expected to approve of their actions.
Figure 1 models the emotional context surrounding the four types of interventions. Emotions elicited by attempting to calm the victim and resolve the problem were virtually identical, as revealed by centroids calculated for each network and the individual graphs. Both conditions showed the strongest links between proud and relieved. Excited and grateful also figured prominently. Adolescents explained that they felt grateful that victims had allowed themselves to be influenced in positive ways. In contrast, amplifying anger yielded strong links between guilty, worried, ashamed, angry, and sad-disappointed. Adolescents explained they were disappointed in themselves. Feeling cold-hearted was linked to ashamed in the amplifying anger condition. Third-party revenge reflected its anti- and pro-social nature with links between proud, relieved, angry, guilty, and sad-disappointed. Angry was strongly connected to proud in the revenge condition. T-tests showed that amplification and revenge differed significantly from each other and the calming and resolving conditions (all ps < .01).
We next examined connections between three moral emotions (pride, guilt, shame) and the three appraisal measures. Overall, proud showed especially strong connections to being helpful and a good friend. Some adolescents described avenging and reconciling responses as “part of being a friend.” Centroids for each network are shown in Figure 2, surrounded by confidence intervals (dotted lines). Those for calming and resolving the problem were situated higher on the positive dimension than those for revenge and anger amplification, and the revenge network was situated higher than that for anger amplification. The subjective meanings of adolescent’s reactions were further examined with subtractive networks that identify nonshared elements of the models associated with each condition. We conclude that ENA is well-suited for identifying patterns of connections between different facets of adolescents’ subjective experience.

Authors

• Karin Frey, University of Washington
  Presenting Author

• Adaurennaya C. Onyewuenyi, The College of New Jersey
  Non-Presenting Author

• Zoe Higheagle Strong, Washington State University
  Non-Presenting Author