New Research Says Robots Could Be Used to Teach Children Good Habits

Table of Links

• Abstract and I. Introduction
• II. The Casa Paradigm and Gamification
• III. Land of Hands: The Game
• IV. Methods
• V. User Study
• VI. Results
• VII. Discussion

  • VIII. Conclusion

  • IX. Acknowledgments and References

Abstract— This paper presents the design, implementation, and evaluation of a novel collaborative educational game titled “Land of Hands”, involving children and a customized social robot that we designed (HakshE). Through this gaming platform, we aim to teach proper hand hygiene practises to children and explore the extent of interactions that take place between a pro-social robot and children in such a setting. We blended gamification with Computers as Social Actors (CASA) paradigm to model the robot as a social actor or a fellow player in the game. The game was developed using Godot’s 2D engine and Alice 3. In this study, 32 participants played the game online through a video teleconferencing platform Zoom. To understand the influence a pro-social robot’s nudges has on children’s interactions, we split our study into two conditions: With-Nudges and Without-Nudges. Detailed analysis of rubrics and video analyses of children’s interactions show that our platform helped children learn good hand hygiene practises. We also found that using a pro-social robot creates enjoyable interactions and greater social engagement between the children and the robot although learning itself wasn’t influenced by the pro-sociality of the robot.

I. INTRODUCTION

Games and social robots captivate and enthrall children. Our study explores the extent to which behaviour change precursors, learning and engagement take place when children and social robots engage in a collaborative game-play around handwashing practises. We first discuss the design and implementation of our novel collaborative educational game titled “Land of Hands” involving children and a customized social robot (HakshE) that we designed. We then present the findings of a user study that we conducted with children aged 6-10 years to evaluate our platform.

Nudge theory postulates that positive and lasting behaviour change such as good handwashing practises could be brought about by repeated nudges that influence the cognitive and affective systems [1]. Facilitating long-term engagements where teachers and parents nudge children towards good hand hygiene practises could be a tedious process that might take several months to achieve. To address this, we propose to use robots as motivational agents that are designed as playmates. Well-designed robots require minimal human involvement and are not perceptible to challenges such as fatigue. Their novelty, especially in developing countries, attracts children towards them enabling them to be used as agents for nudging [2]. The long-term goal of this work is both to build an autonomous social robot that promotes positive behaviour change through nudges and to measure the influence of nudges exerted by the robot on children.

A. Collaborative Games for Positive Reinforcement

Behaviour change may be achieved when children have both declarative knowledge and procedural knowledge about handwashing. The children have to know the steps of handwashing in order to perform it. In our previous research studies, we found that rural and semi-urban school children in India do not have the correct knowledge about hand hygiene steps as prescribed by the World Health Organization (WHO) [3]. Therefore, we decided to blend declarative knowledge and procedural knowledge on hand hygiene into a serious game that the children could play with the social robot. Prior research shows that when robots acted as tutors or peers during gameplay with children, robots were successful in delivering the necessary learning outcomes and encouraged children to remain continuously challenged and motivated [4], [5].

Bartneck et al. created an opportunity for human participants to engage with a robot through a collaborative version of the game, Mastermind [6]. They found that when the robot gave unintelligent advice to human players, then the game was perceived as more difficult and the participants rated the value of the robot’s recommendations to be low. When robots acted intelligently and agreeably, they were perceived as being alive and enjoyable. In a similar study, Xin et al. presented a ‘Sheep and wolf’ mixed-reality game, with Sony’s Aibo robots where they reported that the participants considered the robots as teammates and had enjoyable interactions with them [7]. Thus, we believe that using a social robot to teach good hand hygiene practises to children in a collaborative game-based setting would provide the most enjoyable and positive reinforcement for bringing the behavior change itself.

During the school year 2021-22, many primary schools in India remained closed due to the Covid pandemic. Therefore it was not possible for us to study the influence of our robot’s nudges on children’s hand washing behaviour change. Instead, as a pilot, we conducted this research study to explore the influence of a social robot’s nudges on behaviour precursors, namely children’s learning and engagement with the robot. We will be conducting an offline study during the arXiv:2206.08075v2 [cs.RO] 17 Jun 2022 2022-23 school year as an extension of this work to study the influence on behaviour change.

B. Measuring Social Interactions and Engagement Between Children and Robots

Literature in the field of HRI states that children engage with a robotic system through various verbal and nonverbal communicative channels such as verbal responses, facial expressions, gestures and gaze [8], [9]. Moreover, a robot’s ability to portray pro-social capabilities has been known to positively influence children’s engagement with a robot during long-term interactions [10], [11]. Therefore, to understand how a pro-social robot’s nudges can influence children’s learning and interactions with the robot in a collaborative game-based setting, we measured children’s verbal communication and facial expressions to the robot’s responses.

Through this research study, we aim to answer the following research questions:

• Research Question 1: To what extent does learning about hand hygiene take place in collaborative play between a child and a social robot?

• Research Question 2: To what extent do a pro-social robot’s nudges influence the learning, interaction, and engagement of a child with a robot in a collaborative gameplay setting?