What are children’s beliefs on robots during a “learning by teaching” writing scenario?

Introduction Learning by teaching is a relevant child-robot interaction that could foster learning while increasing the child’s self-esteem and motivation. Beliefs children may have about robots could influence the motivation and have an impact on the learning strategy. Materials and methods During “the school day” in Ecole Polytechnique Fédérale de Lausanne, 55 children along with their teachers interacted with Nao, a humanoid robot, and answered a questionnaire about beliefs they have on the robot. Several explanatory dimensions underlying these beliefs were hypothesized: perception dimension, inferences induced by the scenario, perception of the robot’s autonomy, in order to understand how children perceive robots and make beliefs about them. Results Children are well aware of the robot’s abilities to talk and move. They are convinced the robot had writing difficulties and was progressing in the course of the CoWriter exercise, which is essential for the protégé effect to work. Children also have arbitrary beliefs concerning the robot. For example, they attribute it emotions or cognition. As for the autonomy, it was clear for them that the robot was controlled and programmed by humans. The majority also knew the robot was not alive. Contrary to our hypothesis, the PCA scree plot (Figure 14) suggest the existence of 3 dimensions. These dimensions are better explained by pleasantness, learning process, and skepticism Discussion Even though children were aware of the actual abilities of the robot, mystification still occurred as they believed that the robot had writing difficulties and was learning to write. Further studies need to assess this scenario on children with a wider age range and with neurovelopmental difficulties to better understand these beliefs and their development in typical and atypical children development.