A Child and a Robot Getting Acquainted - Interaction Design for Eliciting Self-Disclosure

In order to facilitate a sustainable long-term interaction between a child and a robot they need to get acquainted with one another. In this paper we discuss the foundation, the rationale, and the evaluation (N = 75) of our design for an autonomous robot conversational partner that engages with Dutch children (8-11 y.o.) in a getting acquainted interaction. The main objective of the robot is to elicit children to self-disclose. Firstly, we discuss five interaction design patterns (IDPs) that proved to be successful in autonomously eliciting and processing self-disclosures. Secondly, we compared two robot behavior profiles. The behavior profiles can be relatively considered as being more and less energetic. We manipulated the movement speed, the speech rate and volume, the use of high/low energy language, waiting time before responding, and the order of high/low energy activities. Results show that the less energetic behavior profile significantly leads to more self-disclosure.