Using automated and fine-grained analysis of pronoun use as indicators of progress in an online collaborative project

Multimodal discourse analysis has been shown to be useful in adding depth to our understanding of the processes of computer supported collaborative learning. We take one of the codes used in a multimodal coding scheme (CPACS), and apply automated data extraction techniques to a large corpus of data aimed at one code in the micro-level (pronouns). The results of this are plotted over time, and patterns of pronoun use identified for further investigation using an in-depth systemic functional linguistics (SFL) approach. Complications concerned with singular and plural second person are discussed, and patterns of pronoun use that indicate movement of the group from one phase of design work to the next are identified. Further refinement of the techniques of automated extraction are required to capture additional patterns noted in the SFL analysis.