Chris Lawson

Chief Scientist

Research statement

I wish to make key contributions to several bodies of research centered on group decision-making in engineering systems. The intent of my research is to develop a methodology that can assess socio-technical systems quantitatively. While it is somewhat straightforward to quantify the behavior of the technical aspects of an engineering system, it is more difficult to assess, in rigorous fashion, the social aspects—specifically in regarding how stakeholders make decisions regarding the design, employment, and operation of such systems. My doctoral research and work experience have led me to the conclusion that a multi-layered approach that integrates empirical investigation and mathematical modeling of stakeholder interactions and behavior is a best approach for this area of research.

From the empirical point of view, I have found that the ability to quantify stakeholder behavior depends greatly on the quality of data collected. However, such data (stakeholder preferences, beliefs, internal decision making rules, and social network information) is difficult to illicit from a real-world system. As part of my ongoing research I have begun developing data collection techniques that utilize a series of interview protocols, surveys, and archival information that have show some success. It is my intent to continue to develop these techniques and hone their efficacy over a large number of case studies.

 In terms of mathematical modeling I have found that mastery of several techniques are useful for quantification of stakeholder behavior in socio-technical systems. These include game theory, social network analysis, and agent-based modeling. Game theory offers a powerful, though improvable tool, for modeling stakeholder decision making and interactions. In my doctoral thesis I found that decision making in committees involves both cooperative and non-cooperative stakeholder behaviors. In order to model this I developed a strategic coalitional bargaining in which reputational effects influence the selection of stakeholder preferences. I wish to continue making contributions to both the application of game theory to committee and organizational decision making, as well as, original theoretical contributions to game theoretic bargaining theory, equilibrium selection, and mechanism design.

 I also find social network analysis to be a powerful tool that can be used to study the interaction of stakeholders within organizations. I believe that there is an opportunity to make original contributions by simultaneously examining the propagation of beliefs, the development of organizational norms, and culture (organizational standards and protocols) within a social network structure. This requires one to first develop an approach that can measure and quantify stakeholder beliefs, decision rule-sets, and the effect of culture on decision-making. One may then model belief propagation and the development of norms via the inclusion of game theoretic interactions on social networks. I am currently working on a paper toward this end. In this paper I first empirically examine the evolution of group norms and belief propagation within four engineering design teams in DOD. I then develop a principal agent model embedded on a social network where group norms act as restrictions on the selection of different game equilibria. Using metrics derived from the empirical analysis I test the model against historical interactions of the four design teams. The initial findings have proven to be promising and there are many possible trajectories for future research.

While closed form axiomatic approaches are often highly desirable, at times it is advantageous to utilize heuristic methods for the study of the social aspects of socio-technical systems. Due to theoretical limitations and issues of complexity, it is often not possible to model organizational decision-making using game theory or social network analysis alone. In this regard, I support the use of agent based modeling in combination with game theoretic and social network models. While one cannot often make generalizabile claims using an agent based model, many insights can be gained which may lead to hypothesis generation or arguments which may be generalizable or at least testable. My research interest lies in the development, integration, and inclusion of agent-based models for theory construction of group and organizational decision-making.