A dynamic research strategy was employed to track the development of mental models of simple DC circuits and reasoning patterns of students learning these concepts for the first time. The medium for research and instruction was a computer tutorial that allowed students to explore their ideas and beliefs as they manipulated simple materials. The study utilized structured observations generated from videotaped data and transcribed analysis of students' verbal commentaries. Verbal protocol analysis (Ericsson and Simon, 1994) provided data for the development of conceptual maps (Dykstra, Boyle, and Monarch, 1992) from which students' conceptual frameworks and mental models were inferred. The sequence and types of changes the models underwent were illustrated by the changes in the conceptual maps as the subjects progressed from a naive understanding towards a more scientific understanding. Both assimilation and accommodation occurred to different degrees: from the acquisition of a single, simple idea (or separation of a single, minor idea) to the development of elaborate and inter-related ideas. In both cases, the process was not simple nor straight forward. Rather, the process is better described as a painful negotiation and renegotiation of conflicting beliefs. The "path" from naive towards expert understanding goes in both directions, with subjects moving backward (toward personal theories) as easily, if not more so, as forward (toward expert understanding). Studying conceptual change in physics learning is necessary in order to develop appropriate instructional materials and strategies that take into account students' preconceptions and how those preconceptions change during instruction. / Ph. D.