| dc.description.abstract | On a daily basis humans interact with an increasing variety of personal electronic devices, ranging from laptops, tablets, smartphones, and e-readers to shared devices such as projected displays and interactive, digital tabletops. An emerging area of study focuses on understanding how these devices can be used together to support collaborative work. Where prior research has shown benefits of devices used individually, there is currently a lack of understanding of how devices should be used in conjunction to optimize a group's performance. In particular, the research presented in this dissertation combines qualitative and quantitative analyses of group work in three empirical studies to link the use of shared and personal devices to changes in group performance and process.
In the first study, participants performed an optimization task with either a single, shared projected display or with the shared, projected display and personal laptops. Analyses of study data indicated that when personal displays were present, group performance was improved for the optimization task ($p = 0.025$). However, personal devices also reduced a group's ability to coordinate ($p = 0.016$). Additionally, when personal devices were present, individuals primarily used those devices instead of dividing time between their laptops and the shared display. To further investigate the support that shared displays provide groups, and in particular, how shared displays might support group work in multi-display settings, a follow-up study was conducted.
The second study investigated how two different types of shared displays supported group work. In particular, shared workspaces, which allowed multiple users to simultaneously interact with shared content, and status displays, which provided awareness of the overall problem state to groups, were investigated. While no significant impact on group performance was observed between the two shared display types, qualitative analysis of groups working in these conditions provided insight into how the displays supported collaborative activities. Shared workspace displays provided a visual reference that aided individuals in grounding communication with their collaborators. On the other hand, status displays enabled the monitoring of a group's overall task progress. Regardless of which display was present, an individual's gaze and body position relative to the shared display supported the synchronization of group activities.
Finally, where the previous two studies identified collaborative activities that were supported by the use of shared and personal displays, the experimental task performed by participants did not explore the transfer of task materials between shared and personal devices or alternative personal and shared devices. The third study addressed these limitations through the adoption of a new experimental task that enabled the exploration of how the manipulation of task artefacts supported collaborative activities, and alternative shared and personal devices in the form of interactive digital tabletops and tablet computers. In particular, the third study compared how personal and shared displays supported sensemaking groups working under three conditions: with shared, digital tables, with shared digital tables plus personal tablets, and with only personal tablets. Quantitative analyses revealed that the presence of the shared, digital tabletop significantly improved a group's ability to perform the sensemaking task ($p = 0.019$). Further, qualitative analyses revealed that the table supported key sensemaking activities: the prioritization of task materials, the ability to compare data, and the formation of group hypotheses.
This dissertation makes four primary contributions to the field of Computer Supported Cooperative Work. First, it identifies cases where the presence of shared and personal displays provide performance benefits to groups, and through qualitative analyses links these performance benefits to group processes. Second, observed uses are grounded in an established process model, and used to identify collaborative activities that are supported by personal and shared devices. Third, equity of participation on shared displays is found to positively correlate ($p = 0.028$), and equity of participation on personal displays is found to negatively correlate ($p = 0.01$) with group performance for sensemaking tasks. Fourth, the method for studying group process and performance based on teamwork and taskwork provides a useful foundation for future studies of collaborative work. | en |
