An ultimate goal of education is to help learners become self-regulated, autonomous learners who can continue to learn beyond formal school education by strategically choosing to use available resources around them. My dissertation approaches this grand question with a specific focus on the topics of choosing to use visual representations strategically and effectively learning with them during algebra problem solving. Visual representations are a type of instructional aids that help learners' sense-making processes during problem solving and learning. I have used classroom experiments, user-centered design, and educational data mining approaches to investigate and support students' learning with visual representations and their learning of strategic use of visual representatios.

In my dissertation, I studied two key questions: 1) how might we support students' effective and efficient algebra learning with visual representations? and 2) how might we support students' self-regulated, strategic use of visual representations during algebra learning? In earlier work, I conducted five studies involving user-centered research and classroom experiments to design and establish novel interactive visual scaffolding called "diagrammatic self-explanation" in the context of intelligent tutoring software for algebra learning. In my later work, I conducted three design and experimental studies to design and evaluate an adaptive metacognitive intervention that supports students' self-regulated use of visual scaffold in the intelligent tutoring software in classroom.

My work with about 20 middle-school teachers and 500 students in the U.S. provides several new contributions in the field of the learning sciences. Among others, my dissertation shows that novel, interactive diagrammatic self-explanation activities for algebra designed with teachers supported effective and efficient learning. My work also shows that a metacognitive intervention designed with middle-school students facilitated students' strategic choices in using visual scaffolding in an interactive learning environment and enhanced both conceptual and procedural learning in early algebra, a challenging dual goal in the field.

176 pages

Thesis Committee:
Vincent Aleven(Chair)
Geoff Kaufman
Amy Ogan
MaRtha W. Alibali (University of Wisconsin-Madison)
Timothy Nokes-Malach (University of Pittsburgh)

Jodi Forlizzi, Head, Human-Computer Interaction Institute
Martial Hebert, Dean, School of Computer Science

Return to: SCS Technical Report Collection
School of Computer Science homepage

This page maintained by reports@cs.cmu.edu