Mental Health Benefits of a Robot-Mediated Emotional Ability Training for Children with Autism: An Exploratory Study

Background: Children with Autism Spectrum Disorder (ASD) have a high prevalence of mental health problems that are linked to reduced emotional abilities. Therefore, interventions that teach emotional abilities are fundamental for their development. However, existing interventions are costly, of difficult access, or inefficient for children with ASD. Furthermore, children with ASD have a preference for sameness and routines that makes technology, and especially robots, an ideal medium to convey interventions that are suitable to their needs.

Objectives: The aim of the present exploratory study is to evaluate whether a robot-mediated emotional ability training is effective in enhancing the emotional ability and the mental health of children with ASD.

Methods: Using a pre-post training design, 12 children with ASD (all boys) aged between 8 and 14 years (M = 10.93; SD = 2.46) undertook a 7 week long emotional ability training mediated by a robot. Sessions took place weekly and lasted 1h each. Children were compared before (T1) and after (T2) the training on their emotional ability and their mental health. Emotional ability was measured through the parent-report measures Emotion Regulation Checklist (ERC; Shields & Cicchetti, 1997), Emotion Regulation Rating Scale (ERRS; Carlson & Wang, 2007), Self-Control Rating Scale (SCRS; Kendall & Wilcox, 1979), and the Alexithymia Questionnaire for Children (Rieffe et al., 2006); as well through a direct measure of children’s use of emotion regulation strategies using the Reactive and Regulation Situation Tasks (Carthy et al., 2010). Mental health was measured through the parent-report measures Children Behavior Checklist (CBCL; Achenbach & Rescorla, 2001), Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997), and the Social Responsiveness Scale-2 (SRS-2; Constantino, 2002).

Results: It was found that regarding emotional ability, children’s use of emotion regulation strategies in the Reactive and Regulation Situation Task, improved significantly after the training (t(10) = 2.81, p < .01) but no significant improvements were found on the parent-reported measures (ERC: t(10) = 0.43, p = .34; SCRS: t(10) = 1.26, p = .12), except for a marginally significant effect on children’s emotional control (ERRS: t(10) = 1.79, p = .05). Regarding mental health, the training significantly reduced internalizing problems (CBCL: t(11) = 1.91, p < .05; SDQ: t(11) = 3.19, p < .01) and autism-related symptomatology (SRS-2: t(11) = 3.24, p < .01), but did not have an effect on externalizing problems (CBCL: t(11) = 0.41, p = 34; SDQ: t(11) = 3.13, p = 07).

Conclusions: Overall, the results of the present study are to be interpreted cautiously, they provide restricted evidence of positive effects of the robot-mediated emotional ability training in children’s use of adaptive emotional abilities and in mental health issues such as depressive symptomatology and anxiety as well as autism-related social communication difficulties. This exploratory study contributes to the research progress in the domain of robot-mediated interventions for children with ASD.