Affective neuroscience/computing approaches in evaluating an interactive educational tool to counteract bullying

TitleAffective neuroscience/computing approaches in evaluating an interactive educational tool to counteract bullying
Publication TypeConference Paper
Year of Publication2016
AuthorsBillis, Antonis, Styliadis Charalampos, Baka Maria, Arfaras Giorgos, and Bamidis Panagiotis
Conference NameSAN2016 Meeting
Date Published08/2016
Conference LocationCorfu, Greece
KeywordsActivity tracker, adolescence, Affect button, Affective Neuroscience, Bullying, Emotiv, facial expressions, Responsible research and innovation, Students, teachers

School bullying is the aggressive behavior among school aged children and adolescents with the intent to hurt. It is expressed in a pattern of unwanted, and unpleasant actions of no justification (i.e. spreading malicious rumors, making threats, physical or verbal violence, social exclusion) that systematically repeat over long periods of time. School bullying is evidently enjoyed by a powerful person or group that aim to control or harm others within any school community. The most common place for bullying to occur among school grounds is the playground, but can extend to the transit between school and home, or at parties. Overall, bullying has a negative impact on everyone involved. In particular, the negative impact on the psychological and socioemotional development, and further life of the children and adolescents involved, may range from poor physical health, increased depression and/or anxiety, behavioral issues, diminished school performance and dropouts, to even psychotic symptoms, emotional distress, and deteriorating physical health. Given the seriousness of both the impact and incidence of bullying within the social or group context in which it occurs, numerous intervention programs have attempted to educate the members of the school community in an effort to reduce the aforementioned outcomes. The main purpose of this research focuses on exploring the affective responses of adolescents induced by multimedia content featuring school bullying. Also, it aims at examining whether the subjective perception and attitude of adolescents towards bullying phenomena are consistent with the results of objective measurements. To this aim, we employed a cross media interactive educational tool that was created in the context of the project ‘European – AntiBullying Campaign’ (Alevizos, Lagoumintzi, & Salichos, 2015) so as to explain the phenomenon of school bullying, awaken children and adolescents in relation to the emotional state of the people involved in bullying, and present practical solutions to counteract bullying. Our experimental protocol, based on the aforementioned tool, was designed so as to allow students to interact with short affective events (~ 20 seconds) featuring scenes of school bullying from a victim centered perspective and their consequences to both the target and the bully, while at the same time measuring multimodal signals of their implicit and explicit affective responses to these events. The sequence of these affective events was deliberately interrupted for 15 seconds by a black screen with written instructions urging the participants to affectively rate the passively viewed short event. This resulted in an experimental procedure of approximately 12 minutes in duration. The participants were provided with a set of questionnaires, including demographics and the Peer Relations Questionnaire (adapted to Greek) (Tsiolka, 2009), which assesses the relationship of students with their peers in their everyday life. Our experimental setup for the synchronous multimodal recordings during the passive viewing of the interactive material, included an activity tracker (FitBit Charge HR) worn on the participant’s non-dominant wrist so as to measure the heart rate, and the webcam of the laptop presenting the material so as to capture the facial expressions, and categorize them into either happy, sad, angry or surprised, as well as estimate the emotional dimensions (valence, arousal) related to the facial expressions. In addition, we employed Emotiv EPOC, a 14 electrode EEG device, which provided us with the affective metrics of frustration and excitement. Finally, in order for the participants to affectively rate the short events, we employed a visual scale-tool, AffectButton (Broekens & Brinkman, 2013), as an intuitive way to define current emotion. A timestamped vector of three dimensions, arousal, dominance, and valence, was produced as a result. Here, we present the results of a single case; an adolescent female, fourteen years old. Prior to her participation to the experimental procedure, informed consent was given by her parents. The participant’s answers to the questionnaire, profile a person who is intolerant with incidents of school annoyance, and shows high levels of empathy for peers that face similar incidents in their daily life. Our findings regarding heart rate suggest that the participant’s response is positively correlated to scenes involving physical violence and bully confrontation. Emotion arousal, as measured by both the EEG affective metrics, namely the excitement (arousal) response levels, and the facial expressions as processed by the video analysis, shows a similar pattern during scenes of escalating tension, i.e. bully physically hurts its victim. In addition, the subjective rating (via the AffectButton) of school violence scenes, shows the participant’s negative (low valenced) attitude towards such scenes. Therefore, combining objective measurements of emotion and its subjective rating, we suggest that the adolescent’s affective state tuned to bullying cues, is characterized by high levels of arousal and negative values of valence. Overall, multimodal, objective measurements of emotional responses to bullying content seem to be in accordance to the subjective perceptions of the adolescent as indicated by the administered questionnaire. This work presents preliminary results of a single case on the affective evaluation of educational content, which aims to increase students’ awareness about school violence and bullying. Immediate plans are to recruit at least ten adolescents, as well as, equal numbers of other members of the school community (i.e. teachers, parents). This is a first attempt to rate in an affective way official school bullying campaign contents. Its impact is believed to drive and improve designs of similar school campaigns, and provide some useful insights to policy makers so as to better empower the members of the school society to counteract bullying in schools. To conclude, the current study uses principles of affective computing and affective neuroscience, based on the axes of Responsible Research & Innovation (RRI). Public engagement in RRI contributes to fostering more societal relevant and desirable research and innovation outcomes to help in tackling societal challenges. Despite the single case character of the present report, the fact that the presented paradigm aims to drive the attention of relevant stakeholders to more acceptable educational content related to school bullying cannot be under-estimated.