Abstract
Although research on children’s trust in social robots is increasingly growing in popularity, a systematic understanding of the factors which influence children’s trust in robots is lacking. In addition, meta-analyses in child–robot-interaction (cHRI) have yet to be popularly adopted as a method for synthesising results. We therefore conducted a meta-analysis aimed at identifying factors influencing children’s trust in robots. We constructed four meta-analytic models based on 20 identified studies, drawn from an initial pool of 414 papers, as a means of investigating the effect of robot embodiment and behaviour on both social and competency trust. Children’s pro-social attitudes towards social robots were also explored. There was tentative evidence to suggest that more human-like attributes lead to less competency trust in robots. In addition, we found a trend towards the type of measure that was used (subjective or objective) influencing the direction of effects for social trust. The meta-analysis also revealed a tendency towards under-powered designs, as well as variation in the methods and measures used to define trust. Nonetheless, we demonstrate that it is still possible to perform rigorous analyses despite these challenges. We also provide concrete methodological recommendations for future research, such as simplifying experimental designs, conducting a priori power analyses and clearer statistical reporting.
Highlights
As human–robot-interaction (HRI) research continues to grow in popularity, increasing focus is being placed towards children as a significant user group
Models of interpersonal trust have been described in [81] where trust is defined as “an expectancy held by an individual or a group
With this research we aim to set a common ground for children’s trust in social robots by focusing on five main contributions, (i) defining competency and social trust and their role in cHRI, (ii) distinguishing between social trust and other pro-social attitudes such as liking and friendship, (iii) analysis of robot-related factors that may influence both children’s social and competency trust in social robots, (iv) identification of different methods used to measure children’s trust in robots and the interpretation of them by the researchers, and (v) insights into how children interact with and perceive social robots, developing a deeper understanding of how trust may potentially relate to other high-level outcomes such as learning and engagement
Summary
As human–robot-interaction (HRI) research continues to grow in popularity, increasing focus is being placed towards children as a significant user group. Rebecca Stower and Natalia Calvo-Barajas have contributed to this study. Multiple facets of trust can be defined. Several models conceptualising trust in organisations have been proposed (e.g., [70,71]). Models of interpersonal trust have been described in [81] where trust is defined as “an expectancy held by an individual or a group
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
