The Impact of Negative Relations on Performance and Satisfaction in Group Work Ling Xia 1 , N. Sadat Shami 2 , Y. Connie Yuan 1,2 , Geri Gay 1,2 1 Department of Communication 2 Information Science Program Cornell University {lx33, ns293, yy239, gkg1}@cornell.edu Abstract This paper reports on a case study conducted to guide the design and implementation of a group support system. Network analysis of student self-reported interactions was conducted to obtain a better understanding of social relations, which are critical in informing the design of systems that support sociability. Building on previous research on group dynamics, the study examined the impact of negative relations and frequency of communication on performance and satisfaction among 13 groups of students. Results show that group members disliked by others were less likely to perform well; group members were less likely to feel satisfied with the group process when they disliked others; and frequent communication with others could make a person more likeable and help him/her perform better. We conclude by providing suggestions about how to incorporate findings from network analysis into the design of communication technologies to support group work. 1. Introduction Groups have been widely recognized as the key organizing unit in contemporary organizations [1, 2]. Group work promises wider access to new information and a greater pool of diverse expertise. However, not all groups collaborate effectively [32]. Advances in information and communication technology can help support the collaboration process as well as help manage its complexity. Scholars and practitioners have studied and experimented with different types of systems to support group work, including for example, expert systems, group decision support systems etc. [e.g., 34]. Yet despite significant investments in these information and communication technologies, many of these systems have encountered social and technical barriers [6]. Preece [35], and Barab, Kling & Gray [5] observed that the design of many existing group collaboration support systems focused too much on issues related to the usability of the technology rather than the support of social interactions. As a result, these systems serve as mere platforms for information exchange, and do not necessarily support the development of social interactions and common values, that are instrumental for collaborative work [35]. Researchers need to understand much more clearly the social processes involved in online collaboration among groups. Requirement analysis for designing group support systems intended to improve social relations can greatly benefit by embedding social network tools in them. Most social network tools have been used to identify positive network structures such as advice, friendship, or communication networks. Using tools to identify negative or adversarial networks to support group work has remained a relatively unexplored area in Information Systems research. Previous research shows that when given a choice between working with a competent person and a likeable person, people tend to prefer a congenial colleague over a more capable one [13]. This preference for working with a highly liked individual over a highly skilled individual has important implications for group dynamics that has been frequently overlooked in empirical research. Social networking tools make it possible to describe interaction patterns among group members and to assess empirically the degree of collaboration among them. This feature can also be used to offer a window into negative group dynamics. For example, social networking tools that can be integrated into group work support systems can provide automatic measures of communication frequency [20], as well as analysis of message content [28]. While these tools alone cannot solve negative group dynamics; through easy and early identification of bottlenecks, they can ameliorate negative dynamics. In this case study, we were interested in examining how frequency of communication, performance and satisfaction relate to negative network structures, and exploring implications of the results for the design of computer supported collaborative systems. Proceedings of the 40th Hawaii International Conference on System Sciences - 2007 1 © 1530-1605/07 $20.00 2007 IEEE Proceedings of the 40th Annual Hawaii International Conference on System Sciences (HICSS'07) 0-7695-2755-8/07 $20.00 © 2007