HCC: Small: Learning-by-Explaining to a Virtual Human

University Of Southern California, Los Angeles CA

Investigators

Abstract

The primary goal of this project is to better understand the cognitive and social factors in a learning paradigm called learning-by-explaining and to build a virtual training partner to promote better learning in this paradigm. Learning-by-explaining is an effective learning technique used by human tutors that has yet to be exploited by the intelligent tutoring system community. In this technique, students are encouraged to explain a concept either to another or themselves. Decades of research shows that generating such explanations can lead to deep understanding of the learning material, and that these learning effects are particularly strong when the explanations are delivered in a social context (i.e., explaining to a peer or tutor), as opposed to explaining to oneself. These effects have even been observed when the "other" is a computer generated character. There are competing views on why learning-by-explaining works. Cognitive theories emphasize how the act of generating an explanation helps student recognize gaps and conflicts in their mental models and creates opportunities for mental model revision and that learning partners facilitate this process by identify missing knowledge and prompting for further clarification. In contrast, social theories argue that the presence and behavior of the explainee motivates learners to invest more effort into the learning tasks, resulting in learning gains. This project aims to gain better understanding of the interplay between the cognitive and socio-relational feedback, how they impact the learning-by-explaining process and how to build an explainee agent that facilitates learning-by-explaining. The project seeks to answer two research questions: 1) Will cognitive feedback or socio-relational feedback facilitate learning-by-explaining? 2) How to build an effective virtual training partner "a virtual explainee" in the learning-by-explaining context? This project will inform the development of next generation intelligent tutoring systems and has the potential to significantly enhance basic understanding of the design of human centered computing. It explores factors related to effective multi-modal interfaces and helps to identify crucial factors that impact social impressions and effective interaction which may facilitate a new generation of more human-centric approaches to human learning. The project will also support the research activities of underrepresented groups (the senior researcher is a woman and the work will support the training of an intern from a HCBU-MI). It will enhance infrastructure for research and education by making advanced research tools and corpora freely available to the research community. These tools provide a novel method to study and enhance the effectiveness of computer-mediated and human-computer interaction, allowing the experimental manipulation of key mediating factors in such interactions. The work will advance discovery and understanding while promoting teaching and learning as it will be performed within the context of USC's Centers for Creative Technologies, a university-affiliated federal laboratory with a core mission to develop and disseminate advanced virtual reality training technology with an extensive track record in transitioning technology into the classroom.