EAGER: Cyberlearning with Co-Robotic Teachable Agents

点击翻译按钮获取中文摘要

Cyberlearning technologies that incorporate robust personalized support and motivational engagement with co-robotic teachable agents present compelling high risk-high payoff opportunities for advancement and broad dissemination of collaborative-robotics STEM education. The project will create robotic teachable agent technologies that provide personalized social, affective, and cognitive support to improve students' STEM learning and motivation. The project will engage at least 10 teachers and 250 middle school students including many underrepresented minority middle school students in the study of collaborative robotics and STEM cyberlearning. The researchers will involve students and teachers in the design and integration of ChalkTalk (a Mixed Reality gesture-based storytelling and simulation tool-kit) with Robotic-Teachable Agents for middle school Geometry (R-TAG) to create an open-source tool-kit for formal and informal STEM learning. The researchers will conduct empirical studies to develop rich models of student cognitive and social interactions, and improve understanding of how to design personalized cognitive, social, and affective support using physical and virtual teachable robotic agents. The project will create a low-cost opportunity for students and teachers to engage in collaborative robotics and STEM cyberlearning. The project will contribute to development and testing of education strategies for broadening participation of students from groups underrepresented in education pathways to careers in robotics. This research will advance the state-of-the-art of co-robotic cyberlearning, engaging hundreds of underrepresented minority middle school students in personalized Ubiquitous Cyberlearning with Collaborative Robotic Experiences (UCCRE). The project will (1) investigate how UCCRE can develop perceptual and cognitive support through collaborative embodied personalized interaction that generates exploratory insight, adaptive feedback, and self-explanation - going beyond typical tutoring support (offering hints, instructional explanations); (2) investigate how UCCRE can develop social support, through embodied personalized interaction with robotic teachable agents that engage and motivate students and (3) evaluate UCCRE through a series of pilot studies using design-based research methods, affective and physiological measures, and Natural Language Processing (NLP) tools. The researchers will use these methods, measures, and tools to sense and adapt to learners' affective states in ways that promote learning and creativity. Students and teachers will iteratively advance the capacity of UCCRE to provide personalized adaptive cognitive and social support, and facilitate collaborative embodied interactions between robotic teachable agents, students and teachers, within formal classroom and informal after-school settings. The findings on participatory design of ubiquitous co-robotics will likely generalize across domains and improve students' interest and motivation in learning robotics.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.