Analyzing the effect of location, communication regime, and demographic faultline on team cohesion

Abstract

Purpose Building on the Lau and Murnighan’s theory of fault line strength, Flache and Mäs (2008b) proposed a computational opinion dynamics model to explore the effect of demographic fault line strength on team cohesion. This study aims to extend the Flache–Mäs (FM) model to incorporate geographical location and the dyadic communication regime in opinion formation process. More specifically, we make spatially proximate agents more likely to interact with each other in the dyadic communication regime. Our results show that when agents update their opinion after each pairwise encounter, opinion polarization is lower at steady state compared to when they update their opinion after interacting with all agents. In addition, if nearby agents are more likely to interact with each other, we see greater polarization compared to the FM model with the dyadic communication regime. An immediate policy implication of this result is that organizational managers should design work space in a way that encourage wider communications between members of a team and avoid geographically local communication. Design/methodology/approach We introduce our computational models to study the effect of location and the dyadic communication regime on team performance (as measured by agents’ opinions on various work-related issues) in the presence of a strong demographic fault line. Our models are extensions of the FM model. For clarification purposes, first we describe the FM model and then elaborate our extensions. Findings The most important finding of this paper is that the timing of interactions plays an important role in steady state of opinion space in a given population. The reason can be traced to the path-dependent nature of social systems, in which initial adopters of a certain opinion or an ideology can significantly change the final configuration of a population. For example, if an early adopter of a given work-related issue in an organization has an extremely positive view toward that issue, and s/he interacts with nearby employees who have similar demographic attributes, we would expect to find an extreme opinion cluster with respect to that issue after a while. However, depending on factors that affect the timing of interaction between individuals, we would expect different outcome in the same organization. If, for instance, more extreme people are more likely to interact, the results would be different compared to when moderate agents are more likely to interact. Originality/value One immediate policy implication of the results of this paper is that organizational managers should design work space in a way that encourage wider communications between members of a team and avoid geographically local communication, if they are to temper the negative effect of a strong demographic fault line. However, they should be cautious and take other related findings into account to avoid undesirable outcomes. For example, according to Flache and Mäss’s results, managers can also initially encourage discussion within demographically homogenous groups and avoid controversial work-related issues. In addition, previous studies showed that more contacts between agents may increase opinion polarization. Our results provide no evidence for more complex and modern organizational designs where individuals or teams do not have a fixed location or stable geographical pattern. For instance, in a modern car manufacturing shop floor, it is possible that workers have to move with cars, or operational engineers have to move between different sections and places. Furthermore, there may be a flexible and dynamic work schedule for workers such that they share a same work station but in different time, which requires a more complex model than what we presented in this paper. In this sense, the geographical setting analyzed in this paper should not be generalized to all organizations or companies. We also have no evidence about other critical factors that might affect the communication and activation regime of individuals. For example, one could imagine a case that workers with the same level of skill in a specific work-related issue are more likely to interact with each other. Moreover, some specific organizational structures could impose additional restrictions on who can/should interact with whom.