Carole R. Beal, University of Southern California
Ronald Stevens, University of California-Los Angeles
Abstract: The goal of this project is to model student learning trajectories as they solve scientific problems in the context of an on-line, multimedia system, IMMEX developed by co-P.I. Ron Stevens (UCLA) and to deploy interventions to support the transition from ineffective to strategic problem solving. IMMEX exercises present the student with authentic scenarios, e.g., an unidentified chemical has spilled after an earthquake, and the student must order specific tests in an effort to identify what the substance is and to learn if it is potentially hazardous. Student's actions within the IMMEX system indicate that many students adopt ineffective strategies, and that strategy is related to student gender. To increase our ability to predict strategic behavior, we have integrated an on-line assessment of student motivation into the IMMEX system. Approximately 900 students have now completed both the motivation assessment and a set of IMMEX cases, and data are being used to train the student model network. We have also developed a pedagogical model to guide the deployment of interventions to support strategic efficiency in relation to student motivation and gender. Initial tests of the model focus on interventions delivered via text messages, with transition to delivery via animated agents beginning in Year 2.