A Trustworthy Robot Buddy for Primary School Children

Abstract

Social robots hold potential for supporting children’s well-being in classrooms. However, it is unclear which robot features add to a trustworthy relationship between a child and a robot and whether social robots are just as able to reduce stress as traditional interventions, such as listening to classical music. We set up two experiments wherein children interacted with a robot in a real-life school environment. Our main results show that regardless of the robotic features tested (intonation, male/female voice, and humor) most children tend to trust a robot during their first interaction. Adding humor to the robots’ dialogue seems to have a negative impact on children’s trust, especially for girls and children without prior experience with robots. In comparing a classical music session with a social robot interaction, we found no significant differences. Both interventions were able to lower the stress levels of children, however, not significantly. Our results show the potential for robots to build trustworthy interactions with children and to lower children’s stress levels. Considering these results, we believe that social robots provide a new tool for children to make their feelings explicit, thereby enabling children to share negative experiences (such as bullying) which would otherwise stay unnoticed.

A systematic review on cross-culture, humor and empathy dimensions in conversational chatbots: the case of second language acquisition

“… N/A TORMES can detect the learner's physiological condition and redirect negative affective states, such as anxiety, into good affective states, such as relaxation Shi et al. (2020 Empathy chatbot To propose an empathy-generating L2 learning chatbot based on transfer learning N/A Construct and deploy an English Learning chatbot based on the principle of explainable artificial intelligence using ontology graph (OG) and transfer learning Smakman et al. (2022) The SAMBuddy Storytelling Cuddle To determine whether humor, intonation, and gender influence children's trust in social robots 115 children The results show that robots might be able to build trusting relationships with children and help them feel less stressed Taguchi et al.…”

Social robotics in music education: A systematic review

“…As for their level of expertise, children were divided into three different groups by learning stages (number of years studying piano) following the piano teachers' suggestion, namely, beginners (less than 2 years, n = 11), developing musicians (between 2 and 4 years, n = 10), and advanced performers (over 4 years, n = 10). In turn, the experience described by Smakman et al (2022) included two experiments about confidence between 4 and 6 years and stress between 3 and 6 years.…”

“…Regarding gender, no rules are repeated for this variable. In Taheri et al ( 2019), all participants were boys; in Taheri et al (2021), one girl and three boys took part; the study performed by Song et al (2021) involved 15 girls and 16 boys, and that of Smakman et al (2022), 26 boys and 29 girls in the experiment about confidence, and 30 boys and 30 girls in the experiment about stress.…”

“…In general, 50% of experiences utilized NAO H-21 (Taheri et al, 2019(Taheri et al, , 2021 while 25% resorted to SAMBuddy Storytelling Cuddle, a robot resembling a cuddly toy full of very basic hardware components (Smakman et al, 2022). Only in one case was a robot created ad hoc for the study; Song et al (2021) presented a robot with two roles: evaluative ("categorical specific language, praising effort, with a slow, constant tone, and a calm facial expression, centering on practice, moving little, and dressing formally (that is, shirt)"); and non-evaluative ("indirect abstract language, praising talent, with a fast active tone and a funny facial expression, moving a lot and dressing informally (i.e., striped jersey)").…”

“…Nonetheless, the so-called "social robotics" has barely been implemented in the educational context as an innovative resource for the development of teaching-learning processes contextualized in today's society (Smakman et al, 2022). We must bear in mind that the latter is characterized as being technological, which highlights the need to address the exploitation of any possibility related to digital technology (Roig-Vila et al, 2021;Alonso Ruiz et al, 2022;Cabezas-González et al, 2022).…”