Design, development, and evaluation of child-robot interaction aimed at enhancing the development of computational thinking in preschool children.

Abstract

Computational thinking (CT) is a skill that enables individuals to formulate problems in a way that can be solved by computers. Interest in CT spans both academia, which annually produces a growing body of research on the topic, and educational settings that are integrating CT into curricula worldwide. While most educational initiatives focus on primary and secondary education, CT can also be effectively introduced at the preschool level, such as through the use of educational robots. This thesis compiles a series of studies that explore various aspects of teaching computational thinking at kindergarten level. We examine the current state of the art by investigating tools, activities, and CT evaluation methods, and then narrow our focus to the concept of conditionals— a programming concept that goes beyond simple sequencing and enables the construction of more complex algorithms. We evaluate current approaches to teaching conditionals, expand the programming language of Robotito, an educational robot developed in Uruguay, to include support for conditionals, and conduct a field study using the robot. Our findings not only highlight the potential of child-robot interaction in fostering early CT skills but also offer concrete developments, as well as design and evaluation methods, for future educational robotics initiatives aimed at preschoolers.