Science of Collaboratories
A large number of academic endeavors today involve people from many different locations. Probably the best known of these is the ATLAS project at CERN where physicists from all over the world collaborate in the building of and running the Large Hadron Collider. We have studied a large number of collaboratories, looking in particular for what makes them succeed or fail. We are collecting some basic information on these collaboratories so that we can compare and contrast various kinds, purposes, configurations, etc. We have also been working on putting the ideas together in a theory, the Theory of Remote Scientific Collaboration (TORSC), combining what we know from the literature with what we have seen in the collaboratories we have studied. In addition, we are turning the theory into an interview script to collect data on all the factors we think are important for success, and going back to these collaboratories to determine the veracity of the factors, their weights and interactions. We have also developed and are about to deploy an on-line version of that interview script so that we can collect a wider set of data more easily, both for theory development and to proactively help those in the collaboratories to ensure success.
Laboratory Studies of Partially Distributed Teams
We have turned to laboratory studies to examine some group communication and trust phenomena in more controlled settings. For example, some of our current work is looking at groups of 10 people, some of whom are collocated and some in rooms by themselves. The 10 people play a “serious game” which has them buy and sell parts to each other and form coalitions, communicating with each other through a fast email system. As it turns out, the collocated people form an in-group, paying attention to each other and ignoring those remote. Those who are remote form their own in-group because other remotes are the only ones who will pay attention to them. We have completed three large studies like this (studying how perceived scarcity is handled in these in- and out-groups, and seeing if once being remote and now collocated you have a more equitable strategy as if ‘walking in the shoes’) and are in the middle of running a fourth in which we compare strangers with those who have a previous strong relationship (i.e. belonging to the same fraternity or sorority or sports club) attend to their buddies even though they are remote. This work takes place in the UCI InformatiLab on the 6th floor of Bren Hall.
Agent-Based Modeling of Group Behavior
Running these laboratory studies takes a long time and is very costly. There are many more phenomena we want to explore than time or resources allow. Consequently, we have begun to model group phenomena using agent-based models. Using NetLogo, we program the agents to act as players in our laboratory study serious game, putting rules in their “heads” about who to pay attention to, how to set prices, how much effort is involved in communicating, etc. When we have been successful in modeling the data we have from the laboratory, we can then start varying things (like loyalties, sizes of groups, etc.) and see what happens. If we find a phenomenon of interest, we will then return to the laboratory with the predictions from the model to test with real people.
Social Ergonomics
Just as there are guidelines for physical ergonomics (e.g. how much you can lift in various positions) and cognitive ergonomics (e.g. how much is lost when you multi-task), there is also a set of social phenomena that can be turned into social ergonomics. For example, the norms we have about how closely we will stand near another person depends on our relationship with that person, whether they be strangers, acquaintances, or intimates (called Proxemics). When we communicate with others through video conferencing, for example, these norms come into play in odd ways. For example, if the image on the screen is very small, we are inclined to speak up, as if they are far away. We are now collecting these rules, and recommending to people how to set up their communication with remote people so the affordances of apparent physical distances, among other things, align with the their intentions.
Telemedicine
Some people practicing telemedicine are using a clever combination of video conferencing and robotics. The remote physician’s head and neck are transmitted to the robot’s screen at head height level, with the “body” on motorized wheels that are under the remote physician’s control. The screen has a mounted camera on it, which is moved with the screen to control what the remote physician sees, but also signaling to the patient what he/she is looking at. Some remote physicians use their “presence” in clever ways, fitting a number of the social ergonomic conventions mentioned above. They stand the appropriate distance from the person they are conversing with, they “move their heads” to talk to the person with near eye-contact, they even talk about “striding into the room.” We have made a connection with a physician at Children’s Hospital, locally, who uses these ‘robots’ regularly, and will interview him and others about their control and the recipients’ perceptions.