Automated adaptive support for peer tutoring

Collaborative activities have been shown to be beneficial, provided that students exhibit certain positive behaviors. Unfortunately, these behaviors rarely occur spontaneously. Adaptive collaborative learning support (ACLS), where an intelligent system assesses student collaboration as it occurs and provides assistance when necessary, is a promising area of research that can help scaffold student collaboration. Little is known about how to build these adaptive systems and what effects they might have on collaboration and domain learning. In this dissertation research, I first augmented an existing intelligent tutoring system with a peer tutoring activity and then iteratively designed, built, and evaluated adaptive support for the activity. This dissertation focuses on two broad research questions: (1) Where and how can intelligent tutoring approaches be applied to the development of ACLS, and (2) Are there benefits to using existing domain models developed as part of individual intelligent tutoring systems in ACLS? I began by implementing a learning environment for peer tutoring as an addition to a successful intelligent tutoring system, the Cognitive Tutor Algebra, and evaluating the benefits of peer tutoring without adaptive support (Phase 1). I then added adaptive support for peer tutors in giving tutees correct help, and discovered that while peer tutors benefit from reflecting on their partner's errors, they need additional support in giving tutees conceptual help ( Phase 2). I designed and implemented adaptive help-giving support for peer tutors, and found positive effects of this support on interaction in a classroom study (Phase 3), and on learning in a lab study (Phase 4). In order to conduct these phases of research this dissertation has made two technical advances: the development of an architecture for extending intelligent tutors for collaboration (Development 1) and the improvement of automated assessment of peer tutor chat (Development 3). This dissertation has also explored potential designs for adaptive support that go beyond traditional intelligent tutoring paradigms ( Development 2). This work makes both technological and learning sciences contributions. The technological contributions involve demonstrating how individual intelligent tutoring approaches can be used to model collaboration, and what role intelligent tutoring components can play in collaborative models. For example, I have shown that the automated classification of peer tutor behaviors can be improved using problem-solving features, and that collaborative skills can be traced in the same way as problem-solving skills. This work makes learning sciences contributions by increasing understanding of the effects of adaptive support on student collaboration and learning. In two studies I have demonstrated that adaptive support, compared to fixed support controls, improves the quality of the help peer tutors give and improves their domain learning. As part of this work, I add to understanding of the cognitive and motivational mechanisms by which different types of adaptive support impact student collaboration. Overall, this dissertation demonstrates that adaptive collaborative learning support is a promising research direction for improving collaboration quality and domain learning.