Exploring a Text-Mining Approach for the Analysis of Computer Collaborative Data From a Design-Based Research Project

• In Event: Characterizing How Students Learn Through Collaborative Activity

Abstract

Researchers typically gather far more data than they can analyze, and while potentially robust, qualitative analyses pose limitations for generating fast, reliable analyses of these amounts of data. We explore a computational approach to analyze learning data of pre-service teachers who collaboratively created video-based artifacts. Our results indicate that when pre-service teachers refine their artifacts after discussing with their peers, they pay more attention to student learning in the video, and thus produce artifacts that are more closely aligned with expert artifacts. We argue that computational techniques are replicable and reliable, and can handle greater amounts of data in a timely manner, which is advantageous during rapid implementation cycles, such as in a design-based research project.

Authors

• Alejandro Andrade, Indiana University - Bloomington

• Asmalina Saleh, Indiana University - Bloomington

• Andrea M. Rehak, Indiana University

• Andrea Sarah Gomoll, Indiana University - Bloomington

• Joshua Adam Danish, Indiana University

• Cindy E. Hmelo-Silver, Indiana University