Developing a Coding Framework to Analyze Student-to-Student Reasoning Based on Mental Models Theory

The use of active learning pedagogies, as well as research into their effectiveness, has increased greatly the past few decades. These pedagogies typically depend on student-to-student interactions to facilitate learning. Video recordings of student interactions provide excellent observational data from which to study the dynamics of these pedagogies in a naturalistic setting. However, these data are typically voluminous, include many potential features to follow, and as such make analysis difficult. One way to decrease the difficulty in analysis is to use a robust coding framework. This study develops such a coding framework using a well-established Mental Models theory of reasoning as a theoretical lens. Each element within the coding framework is analogous to an element in the mental models theory. This coding framework was applied to video recorded data of six student teams reviewing a peer team's prototype design in a classroom setting. The coding resulted in 567 transcription segments of which 68% related to the prototype review. All elements of the mental models theory are evident and code-able in the data and the general structure of the verbalized reasoning is identified. A rich description of the verbalized reasoning is provided. Furthermore, this reasoning structure appears constant across changes in student engagement and interaction purposes. As such, the identified structure of student reasoning, based on the mental models theory, provides a robust coding framework.