Exploring and Supporting Today's Collaborative Writing

Collaborative writing has become increasingly common and complex. Many researchers in Human Computer Interaction (HCI) have conducted extensive research on this topic for the last 25 years. Technologies were built and tested in the laboratory. I continue this stream of work by revisiting computer-supported collaborative writing within today’s context. Today, many commercially available word processors have built-in features to support collaboration, such as supporting synchronous editing and archiving editing history. Now that these features are available outside laboratories, how are people actually using these new capabilities in the wild? In my dissertation project, I aim to explore this question using both quantitative and qualitative research methods. In addition, to analyze the collaborative writing practices utilizing the detailed behavioral data traces, I have built an information visualization system, DocuViz, that complements existing research methods. This dissertation work contributes both a visualization system and the novel research approach of using these visualizations to study collaborative writing. This work also reveals various features and practices that people use to write collaboratively. By quantifying collaboration markers, such as styles of work or participation equity, the statistical analyses reveal the relationships between the practices that people use and the quality of the documents that they produce. Our results suggest that people write collaboratively more often than they used to, and they use various styles of working (e.g., outline then divide and conquer or template) to coordinate their collaborations. A few factors (e.g., power dynamics, privacy, and community’s norms) influence the practices that people choose. Some group behaviors (e.g., having a leader or equal participation) are associated with higher quality outcomes. In summary, I construct a holistic understanding of the users’ perceptions and practices that they developed in adapting to today’s collaborative writing technology. By synthesizing both the findings from my research and previous literature, I extend the existing research framework of collaborative writing, propose behavioral guidelines for users who want to write together better, and generate design implications for system designers. In addition, I offer two visualization tools used in this research that might be useful to authors themselves as well as to future research