Diagram Interaction during Intelligent Tutoring in Geometry: Support for Knowledge Retention and Deep Transfer

Prior research has shown that skilled problem solvers often demonstrate meaningful links between relevant visual and verbal knowledge components, but little is known about how to support novice learners in connecting visual and verbal information during learning or whether such support will improve learning outcomes. This work explored two methods to support meaningful connections between visual and verbal information in an intelligent tutor for geometry: 1) student interaction with diagrams during problem solving, and 2) student explanations that connected diagram features to geometry rules at each problem-solving step. Research was conducted in 10 grade classrooms using an experimental version of the Geometry Cognitive Tutor. Results demonstrated that interaction with diagrams promoted longterm retention of problem-solving skills and supported performance on transfer tasks; diagram-rule explanations did not significantly influence learning. Findings suggest that student focus on relevant visual information should be carefully integrated into problem-solving practice to support deep learning.