EVALUATING THE PEDAGOGICAL VALUE OF USING STUDENT-MANIPULATED MODELS IN A LARGE, UNDERGRADUATE, EARTH SCIENCE CLASS

Student manipulation of physical models promotes student inquiry and discovery while confronting potential misconceptions related to earth science concepts. While this pedagogical technique has been widely employed in K-12 and undergraduate lab classes, it has not been used extensively in large (n > 100) undergraduate classes. Several physical modeling activities were designed for student use within a large introductory earth science course for non-majors. These models illustrated key aspects of concepts associated with seasonal changes in insolation, geologic time, plate tectonics, ancient climates, and other concepts throughout the course. The models were used in classrooms that employed active learning techniques, including the use of student personal electronic data collection devices for answering conceptual questions related to the concepts being modeled. A guided inquiry approach was used, whereby; four person teams were presented materials pertinent to a particular concept and provided general guidelines to help them complete a set of modeling tasks. Students were encouraged to work together within their groups to understand the model's underlying principles before individually answering conceptual questions.

Both quantitative numeric data (pre- and post-test conceptual questions) and qualitative data (student interviews and field observations) were collected to evaluate the effect of using these models. For example, archived data for a solar insolation modeling exercise included 264 student responses to 2-3 pre-modeling conceptual questions, 269 responses to 3-5 post-activity questions and 22 pages of observer notes The student response data were analyzed using SPSS for statistically significant differences between control and model classes. The observational data were analyzed for reoccurring themes that documented student usage of the models. To numerically quantify student understanding, a student gain score was defined. Preliminary quantitative results are mixed, but the data provide useful, illustrative information regarding which models are the most effective and for whom.