Kate Vogt (Dr. Duke Hutchings) Department of Computing Sciences
Intelligence analysts must make sense of large amounts of disparate information in order to evaluate and predict various situations, such as a threat to national security. As they work through the sensemaking process, the field of visual analytics seeks to productively help the analysts comprehend large amounts of textual data with concise visualizations. Analysts often work collaboratively in a co-located setting, but prior to this study, there was no work exploring the collaborative analytic process on a large, high-resolution computer display.
Here, we have the explored the concept of large-display collaborative visual analytics by running an intelligence analysis exercise containing a terrorist plot with pairs of people who had experience working together. Each member of the team had their own keyboard and mouse inputs which facilitated the co-located collaborative sensemaking process on the display. Eight pairs of participants served in the study, during which we collected video and audio recordings, display screenshots, and participant mouse actions and user interface window movements.
We also conducted interviews and completed post-hoc video coding to aid the analysis. Evident in the analysis were five unique strategies which together formed the computer-aided collaborative sensemaking process. We also identified spatial and territorial patterns adopted by the participants on the large display and user roles that developed within the teams. Teams which collaborated most frequently also produced the highest scores for the exercise. Combining all of these findings, we have developed design implications for co-located collaborative computer hardware and software tools.