Advancing Understanding of Collaborative Learning with Data Derived from Video Records Randi A. Engle (University of California, Berkeley)
How exactly one approaches or should approach an analysis of collaborative learning using video recordings depends crucially on one’s theoretical commitments, on the speci c research questions being pursued, and on practical constraints of time, money, and personnel. A video analysis is high quality to the extent that the researcher can make a convincing case that one’s analytic choices and argumentation connecting claims to data were su ciently responsive to these considerations. Issues of reliability and validity of all kinds (internal, convergent, external, and descriptive) apply to video-based data as they do to any other kind of quantitative or qualitative data analysis. Concerns about generalizability of ndings can be countered by explicit attention to the logic of one’s inquiry, one’s approach to collecting records, and an articulation of the processes used
to create explanations and generate claims. As a result, performing analyses with video recordings is frequently an iterative process that involves cycling between the video records themselves, one’s evolving hypotheses and data interpretations, and a variety of intermediate representations for discovering, evaluating, and representing them for oneself and others.