Classrooms are social environments. Our largest lecture halls are the size of many small towns and develop their own intricate social structures that can greatly affect teaching and learning. Social network analysis allows us to rigorously analyze the interactions that make up our learning environments. In particular, we use exponential random graph models to test hypotheses using easy-to-collect student interaction data.
Our poster will describe and demonstrate the methods we are using for data collection, analysis and interpretation. This method is widely applicable to questions about student learning, perceptions, and the use of student societies to help or hinder learning. Our focus in this poster is to show the kinds of questions that can be asked and tested.
For example, we can show that students in a 200-student Biology course are extremely accurate in their perceptions of other student’s knowledge (as judged by exam scores that are otherwise confidential). Previously anecdotal, this support for social hypotheses can inform changing classroom styles, especially in large lectures where the social environment dwarfs infrequent instructor-student interactions. This is just the beginning, and we hope to spread this technique to other researchers in education.