Shape-Factory PublicationsIn-group/Out-group Effects in Distributed Teams: An Experimental Simulation CSCW-2004Nathan Bos, N. Sadat Shami, Judith S. Olson, Arik Cheshin, Ning NanABSTRACT Modern workplaces often bring together virtual teams where some members are collocated, and some participate remotely. We are using a simulation game to study collaborations of 10-person groups, with five collocated members and five isolates (simulated ‘telecommuters’). Individual players in this game buy and sell ‘shapes’ from each other in order to form strings of shapes, where strings represent joint projects, and each individual players’ shapes represent their unique skills. We found that the collocated people formed an in-group, excluding the isolates. But, surprisingly, the isolates also formed an in-group, mainly because the collocated people ignored them and they responded to each other. Traveling blues: The effect of relocation on partially distributed teams CHI-2005 short paperNathan Bos, Judith Olson, Arik Cheshin, Yong-Suk Kim, Ning NanABSTRACT This experimental study looks at how relocation affected the collaboration patterns of partially-distributed work groups. Partially distributed teams have part of their membership together in one location and part joining at a distance. These teams have some characteristics of collocated teams, some of distributed (virtual) teams, and some dynamics that are unique. Previous experiments have shown that these teams are vulnerable to in-groups forming between the collocated and distributed members. In this study we switched the locations of some of the members about halfway through the experiment to see what effect it would have on these ingroups. People who changed from being isolated ‘telecommuters’ to collocators very quickly formed new collaborative relationships. People who were moved out of a collocated room had more trouble adjusting, and tried unsuccessfully to maintain previous ties. Overall, collocation was a more powerful determiner of collaboration patterns than previous relationships. Implications and future research are discussed. An experimental simulation of multi-site software developmentN. Sadat Shami, Nathan Bos, Zach Wright, Susannah Hoch, Kam Yung Kuan,Judy Olson and Gary OlsonABSTRACT Effective communication and coordination across multiple sites is extremely important for global software development. An experimental simula-tion that mimics an interdependent software divi-sion working across multiple locations was designed to study this phenomenon. Six experi-ments were run, each with participants divided into four or five sites depending on the availabil-ity of participants. We found that simulated work-ers at the same sites formed strong in-groups and were able to enlist help from their collocated col-leagues at a much higher rate than from remote colleagues. These strong local in-groups inhibited cross-site collaboration. Remote workers had par-ticular difficulty coordinating work with in-group members. However, certain isolated participants did perform well by being proactive in communi-cation and reminding collaborators of reciprocal relationships. Future uses of this simulation will test different interventions such as instant messag-ing, calendaring, and traveling, which may help overcome the challenges of global software de- velopment and inform the design of future col-laboration technologies. Beyond Being in the Lab: Using Multi-Agent Modeling to Isolate Competing Hypotheses CHI2005 short paperAbstract In studies of virtual teams, it is difficult to determine pure effects of geographic isolation and uneven communication technology. We developed a multi-agent computer model in NetLogo to complement laboratory-based organizational simulations. In the lab, favoritism among collocated team members (collocators) appeared to increase their performance. However, in the computer simulation, when controlled for communication delay, in-group favoritism had a detrimental effect on the performance of collocators. This suggested that the advantage of collocators shown in the lab was due to synchronous communication, not favoritism. The canceling-out effects of in-group bias and communication delay explained why many studies did not see performance difference between collocated and remote team members. The multi-agent modeling in this case proved its value by both clarifying previous laboratory findings and guiding design of future experiments. Presentation at CSCW-2004Presented by Nathan Bos |