Talk Abstract
People construct internal representations, or mental models, of how their environments, other people, and technologies function. In learning sciences research, mental models, or situational models, can play a key role in how we understand the way learners engage with learning activities. Learner perceptions of an activity, including (1) the details of a problem or text, (2) the purpose, form, and function of the activity, (3) the intended audience (e.g. peers, parents, or teacher) of resulting outputs, as well as (4) factors like self-efficacy and motivation, all influence learner behaviors, learning processes and knowledge construction. A similar conceptual use of mental models exists in human-robot interaction research to explain the way people interpret how to interact with robots. When encountering a robot, people's perceptions of (1) the details of the shared activity with the robot, (2) the purpose, form, and function of the activity, (3) the robot's role and function, (4) the robot's sensing, computing, and physical capabilities, (5) their own role and function, (6) and their own strengths and capabilities, all influence how they interact.
My research in robot-assisted learning sits at the intersection of these two fields. In over a decade of work examining student-robot interactions, I've come to recognize there seems to be a unique interaction effect of student mental models of learning activities and their mental model of robot interactions. In this talk, I present a preliminary framework to integrate these two conceptualizations of a mental model. I then present insights from several examples of past work to illustrate and explore the potential of this framework to build theory, inform practices, and generate testable hypotheses in robot-assisted learning. Finally, I make the case that this framework extends beyond learning with robots into the broader field of learning with technology. The talk is designed to be an open space for the presenter and audience to explore and refine these ideas and put these insights to the test.

Biography
Joe Michaelis is an Assistant Professor at the University of Illinois Chicago in Computer Science and Learning Sciences. He has a Ph.D. in Educational Psychology – Learning Sciences from the University of Wisconsin-Madison. After six years as a science educator and FIRST robotics coach in urban middle and high schools, he returned to Madison to pursue research interests in STEM education. His research focuses on designing and assessing educational technologies by utilizing human-centered methods and interaction modeling in HCI, and on interest, motivation and socially mediated learning processes in the learning sciences. In this work, he designs and studies technologies that use social connection-making with learners to support learning and interest through long-term engagement, in ways that seamlessly integrate into existing educational activities in classrooms, informal learning environments, and at home. Joe's current research focuses on designing learning companion robots and social robotics curriculum to support learning and interest in STEM fields and is supported through several NSF grants.

URL: https://www.uiclitlab.com/