Cathleen McGrath

Structures that work: social structure, work structure and coordination ease in geographically distributed teams

Scholars have recently argued for flatter, organic organizational structures that enable workers to deal more effectively with dynamic and uncertain environments. In a correlational study of 33 R&D teams, we find that although this network form is associated with more smooth coordination in collocated teams, the opposite is true for geographically distributed teams. In fact, an informal hierarchical structure was associated with more smooth coordination in distributed teams. These results add to the scant literature on networks in teams and provide insight into important differences in the structure of geographically distributed and collocated teams.

Network Conditions for Organizational Change

Understanding the overall network structure of organizations can help managers to support change. This article describes three different network theories of change, exploring the underlying assumptions and implications of each model. First, the E-I model predicts that cross-departmental friendship ties will help generate positive response to change in organizations by fostering trust and shared identity. The viscosity model predicts that introducing controversial (not clearly good or bad) change into the periphery of an organization and carefully regulating the interaction of innovators and nonadopters provides the best chance that it will diffuse successfully. Finally, the structural leverage theory presents a mathematical model that supports broad diffusion of clearly superior change, informing as many people as possible about the change.

The Effect of Graph Layout on Inference from Social Network Data

Social network analysis uses techniques from graph theory to analyze the structure of relationships among social actors such as individuals or groups. We investigate the effect of the layout of a social network on the inferences drawn by observers about the number of social groupings evident and the centrality of various actors in the network. We conducted an experiment in which eighty subjects provided answers about three drawings. The subjects were not told that the drawings were chosen from five different layouts of the same graph. We found that the layout has a significant effect on their inferences and present some initial results about the way certain Euclidean features will affect perceptions of structural features of the network. There is no “best” layout for a social network; when layouts are designed one must take into account the most important features of the network to be presented as well as the network itself.

With a Little Help from Their Friends: Exploring the Advice Networks of Software Entrepreneurs

This field interview study examined patterns and content of advice sharing networks among 20 software executives to provide a clearer understanding of how advice relationships are established, the types of advice that are shared and the role that relationships play in the support of information exchange and diffusion. Most advice relationships were formed from strong tie relationships, while systematic differences were found among the types of advice sought from advice relationships resulting from strong ties, business ties and weak ties. The preference of software executives for rich communication media supports the importance of establishing trust in advice sharing relationships.

Trust: A Collision of Paradigms

The technological challenges of securing networks are great, as recently witnessed in widespread denial of service and virus attacks. The human reaction to these attacks may be either a loss of trust or a willingness to tolerate increasing risk having weathered one assault. Examining human and computer interaction with a focus on evaluations, the human response to loss of trust is a key part of the search for more secure networks. The success of current efforts to design appropriate security mechanisms depends as much on an understanding of human extensions of trust to computers as it does on an understanding of underlying mathematics. However, the former has not been sufficiently examined.In this work we survey the findings in social psychology and philosophy with respect to trust. We introduce three hypotheses that remain unanswered with respect to the manner in which humans react to computers. We discuss potential design revisions in light of findings from other disciplines. Then we conclude by noting that research which empowers users to be their own security manager may be based on a fundamentally flawed view of human- computer interaction. We close by encouraging designers of computer security systems to examine the humans, which these systems are intended to empower, and recommend that any security system be built on the basis of understanding of human trust provided by the social sciences.

Visualizing Multiple Levels and Dimensions of Social Network Properties

In this chapter, we develop a framework for understanding how multidimensional, multilevel data is most effectively conveyed in social network diagrams. We build on work begun in 1994, with a series of explorations of social network visualization with the theme of helping viewers make accurate judgments about network properties. In contrast to contemporary work on layout aesthetics, we were interested in helping viewers with questions that may lie beyond the mathematical graph to the social group that it represented, and that may not be readily available by inspection, unlike path length or degree. We found that different layouts of the same graph had appreciable effects on viewer’s judgment of the importance of individuals with the network individuals or the number of subgroups. In addition, we found that the use of motion to link alternative views of a network had different effects depending on the layouts used. In this chapter we describe and organize some of these results with respect to the long-term goal of aiding human judgment over complex structured data.