Towards the Design of a Robot Peer-Tutor to Help Children Learn Math Problem-Solving

Abstract

Collaborative learning allows learners to work together as peers to understand the given problem and to formulate different strategies for solving. With the rising popularity of social robots, education researchers have explored the varying roles these technologies can fulfill in interactive learning environments, such as tutors and learning companions. In this paper, we describe our work in investigating a robot’s potential as a peer-tutor for Grade 1 students learning math problem solving. Combining audio-visual cues in presenting addition and subtraction problems, our robot, named Vi, worked with 12 children age 6-8 years old as they try to understand and formulate a solution plan for the given problems, using a dialogue framework that leverages on the mathematical thinking process. Our results showed mixed feedback from children; some found Vi fun to talk to while others felt pressured by Vi’s questions specifically for those participants who are not fluent in the English language.