Interdisciplinary (99) This project is designing, deploying, and evaluating software tools that automatically structure and scaffold undergraduate student interactions within online discussion boards. Based on analysis of archived student contributions to previous discussion topics the software tools recommend those contributions that are useful, relevant or of interest to undergraduates currently participating in discussions of the same topic. Rich assessment techniques are being developed to analyze discussions, and these are based on course ontologies and course corpora that the PI team is creating. Tutorial strategies used include generation of questions and comments, invitations to participate in discussions, and clarifying annotations. Furthermore, student profiles of interest or expertise help link individual students to one another on particular topics. Evaluation studies focus on the effects of different scaffolding interventions on student performance, STEM interest, and retention rates. The intellectual merit of the project lies in two areas: 1) the assessment of how the collaboration and reflection features impact STEM undergraduate student performance and student interest in science and engineering; and 2) the development of innovative technologies for automatic scaffolding and assessment of student contributions. The project's potential for broader impacts derives from its support for peer learning, reflection, and collaborative problem solving. These attributes increase student engagement and performance and can deepen conceptual understanding across a wide range and number of students in classroom settings that are not constrained by enrollment, since they are web delivered.
