The Effects of a Socially Assistive Robot on Engagement and Self-Disclosure to improve Emotion Regulation during One-on-One Child-Youth Coach Conversations: A Case Study

Abstract

Numerous children have difficulties in regulating their emotions and suffer consequences such as the development of mental health issues, or deal with societal challenges like criminal behavior, low personal well-being, and academic struggles. This research investigates the potential for improving children’s emotional regulation by increasing their engagement and self-disclosure during one-on-one conversations with youth coaches in the youth care sector, while using a personalized SAR.

A case-study was done where participants talked to a SAR called ’Buddy’. Buddy did not replace the youth coach during the one-on-one conversations, but served as an ’ice-breaker’ in a Wizard-of-Oz set-up that intended to increase engagement and disclosure about emotion-related experiences of the child to the youth coach. Engagement coding and thematic coding, as well as statistical analyses, evaluations and interviews, and questionnaires were conducted to investigate SAR effectiveness.

Deploying the SAR only led to a slight increase in engagement, and there were mixed results in terms of self-disclosure among participants. The correlations between engagement, self-disclosure, and emotion-related themes were generally weakly positive or moderately negative. Contrary to these results, youth coaches held a very positive view of Buddy’s deployment, foreseeing its future potential. Additionally, participants enjoyed conversing with the SAR, and there was evidence of improved emotional regulation over time based on questionnaire responses.

In the end, we cannot conclude that introducing a SAR in one-on-one conversations between a child and youth coach improves emotional regulation. The limited number of participants in this study limits the applicability and reliability of the findings. Although we cannot draw any conclusions, research does describe positive effects of SAR deployment for ER improvement, and we also found some minor positive results that support the effect. Nevertheless, this research sheds light on the reasons behind resistance to participating in experiments where SARs are deployed, and offers suggestions for future prevention of this resistance. Future studies could investigate the impact of SARs with different approaches in various settings, and further explore engagement, self-disclosure, and ER enhancement.