Research Interests & Projects
Through my research, I seek to develop a novel, multi-scale framework for the study of collaboration that will enable us to better use information and communication technologies (e.g., telephone, videoconferencing, text chat, blogs, etc.) to support communication, collaboration and learning in a range of organizational contexts . “Collaboration” can describe collective efforts ranging in duration from a two-minute discussion to a multi-year project, in size from two to thousands of people, and in magnitude of accomplishment from quickly drawing a sketch to sequencing the human genome. Whatever the scale on any of these dimensions, however, collaborations do not occur in isolation and are not independent of one another. Long-term or high-magnitude projects typically have many smaller-scale components, each with constituent collaborators. People frequently move between these in their day-to-day work.
Nonetheless, most studies to date on technologies to support collaboration have been focused at a single scale – be it that of, for example, large, planned projects or quick, ad-hoc meetings in front of whiteboards. We lack an integrating framework that allows us to understand the relationships between simultaneous collaborations at multiple scales, and their implications for the design and development of information and communication technologies. In the long term, I seek to develop a framework for the study of collaboration that takes these relationships into account.
In my work, I have used both qualitative and quantitative methods, in both field and laboratory settings, to address questions from two perspectives: the social and organizational context of collaboration, and the nature of focus and attention in groups accomplishing specific tasks.
The Social and Organizational Context of Collaboration
In order to understand the conditions under which different types of collaboration are likely to succeed or fail, we must first understand context. I have focused on several types of collaboration in two domains.
Improving Collaboration Technologies for Researchers
Advances in technology combined with the increasing cost and complexity of research have led to strong government interest in “cyberinfrastructure,” the emerging collections of networking, computing and communications resources to support research. Using these technologies to enable novel forms of research is a common goal, but there is little consensus about how to accomplish this. It is difficult to understand what specific tools are needed in different cases, difficult to overcome communication and coordination issues introduced by geographic dispersion of collaborators, and difficult to facilitate adoption of tools to bridge these distances.
In my dissertation research, some of which will appear in a forthcoming article in the Journal of the American Society for Information Science and Technology, I used a combination of mail survey and interview methods to improve our understanding of what motivates and sustains collaboration. Specifically, I illustrated that differences in collaboration behavior among researchers in three disciplines were better explained by the nature of individuals’ work and interests than by their attitudes toward individual vs. collective discovery. This suggests that in designing technologies for these settings, we should focus more closely on supporting day-to-day work practices, rather than trying to overcome perceived barriers such as scientific competition and intellectual property. I have continued work in this area through my involvement with Dr. Barry Wellman (University of Toronto) in an investigation of knowledge dissemination in the Canadian Water Network. I plan to continue by conducting additional survey and interview studies investigating a wider range of factors and settings.
In carrying out my dissertation fieldwork, I also had the opportunity to observe novel uses of existing technologies in research. In a series of papers co-authored with several Michigan colleagues (Dr. Thomas Finholt, Dr. Daniel Horn, and Matthew Bietz) I have argued that the requirements of distributed research groups present significant challenges and opportunities to existing technologies. One major issue, for example, is finding ways to integrate remote participants into large-scale experimental research activities. Based on my experience with NEESgrid, a large, NSF-funded virtual laboratory for the earthquake engineering community, I argued that one way to accomplish this is by involving remote participants in tasks that they may be better suited to complete, such as detecting small changes in an experimental specimen by using a slow frame rate video view. This is one example of many difficulties that can be overcome by extending the capacity of current technologies into new and useful realms.
Organizational Character
Supporting effective communication and collaboration in organizations, particularly those that are geographically distributed, requires a broad understanding of how people in these organizations get information and decide what to do when they encounter uncertainty. To investigate this, Dr. Michael Cohen (University of Michigan) and I examined the annual “regeneration” processes among the staff at a children’s summer camp. A summer camp, characterized by nine months of dormancy and high staff turnover rates, provided us with the ideal setting for exploring a fundamental question of organizational change: How do we recognize an organization at two points in time as “the same” organization? In a paper recently provisionally accepted by Organization Science, we proposed the notion of “organizational character,” a coherent and robust set of traits analogous to individual character, which acts as a guiding frame for both newcomers and veterans in determining how to act. Improving our understanding of these processes allows for better targeting of specific areas for technologies to improve communication and knowledge transfer in all types of organizations, but particularly those who seek to retain the same “character” across temporal or geographic gaps.
Focus and Attention in Small-Group Collaboration
In my work as a postdoctoral fellow on the NECTAR (Network for Effective Collaboration Through Advanced Research) project, I have focused closely on a significant problem with many technologies that I observed in my fieldwork. Namely, these technologies lack adequate support for awareness and interaction between local and remote group members. In collocated contexts, for example, people frequently move fluidly between engaged interaction and coarse-grained awareness of their colleagues’ activity. Technologies such as email or videoconferencing, however, are generally geared toward single modes of communication and do not support these transitions well.
One key problem in this area is the difficulty of providing for adequate awareness of remote colleagues’ activity, while simultaneously respecting individual privacy. I am working with Dr. Carl Gutwin (University of Saskatchewan) on a field study of creative professionals who work in open plan office spaces. Data collected so far have provided valuable insights into how participants make use of subtle environmental cues and strong social norms in maintaining a satisfactory balance between privacy and awareness. These data are currently being used to drive the design, implementation and evaluation of a novel suite of instant messaging tools that will be the focus of future studies.
A related problem is common in "hybrid" groups where a local group is joined by a smaller number of isolated remote participants. This is particularly problematic in “webcast” classroom environments where it is easy for a local group to transition between a formal presentation and discussion, but much harder for remote participants to join in. Remote participants may be experiencing the event on a delayed connection, and often have little sense of who else is present and who is speaking. Moreover, it may be difficult for a local presenter to know that a remote participant wishes to speak and to “call on” that person. With Dr. Ron Baecker (University of Toronto) and Dr. Saul Greenberg (University of Calgary), I am engaged in work to improve awareness and interactivity in ePresence, a webcasting tool developed by Baecker’s group. We have focused on better supporting social process (via two field studies I have supervised), and awareness/interaction technologies. We have already developed a preliminary system, and are currently conducting an evaluation of its effectiveness in a class at the University of Toronto. Future work will include experimental incorporation of additional features, such as freehand gesturing to allow the speaker to naturally “call on” remote participants.
Another key issue in understanding interaction mode transitions is the problem of view control for remote participants “tuned in” via a video link. Action in local events, particularly in times of uncertainty or crisis, frequently unfolds faster and in a broader space than remote participants are able to view or track with existing camera and control technologies. As computer vision and machine learning technologies improve, however, it will be possible to anticipate activities of likely interest to remote participants and use this anticipation to automatically control camera shots and vastly improve the experience of remote participants. With Dr. Ravin Balakrishnan (University of Toronto), I am currently working on a series of experimental studies in which we have used motion capture data to drive automated cameras in a dyadic Lego construction task with a remote helper. We have developed a preliminary system that shows substantial performance benefits in terms of time and accuracy. The first of these studies, in a paper that won honorable mention for best paper at the ACM Conference on Computer-Supported Cooperative Work, demonstrated that users adapt their behavior depending on camera condition and view, which has critical implications for prediction based on that behavior. Future work will include refining our automated control system, and removing constraints on the task scenario to more closely simulate real-world tasks.
Summary
By focusing on collaboration at both of these scales, I have contributed to our understanding of the conditions under which collaboration occurs, the nature of collaboration in distributed organizations, and how to better support shifts in attention and awareness in workgroups. I plan to continue exploring these questions as I move toward my long term goal of a multi-scale framework for understanding collaboration and improving the use of information and communication technologies in organizations.
|
|