Computational models, often in the form of virtual worlds, are used in social and engineering policy domains to address via what-if analysis, how different technologies, decisions and policies influence the performance, effectiveness, flexibility, and survivability of complex social systems. Using computational models, researchers examined a number of critical phenomena: the role of cognition, the role of interaction, the role of task, and the role of chance. Findings in these areas have led to new scientific paradigms and adjusted the way in which we think about social and organizational behavior. For example, some of the earliest work focused on demonstrating the fallacy of the economic model of the rational actor. One of the earliest models, "A Behavioral Theory of the Firm" (Richard Cyert, and Jim March) was used to demonstrate the strength of computational analysis for understanding complex human behavior. In the "Garbage Can Model" (Cohen, March and Olsen, 1972) demonstrated that satisficing behavior and bounded rationality cause most organizational decisions not to be made by resolution. Allen Newell and Herb Simon, demonstrated the value of using knowledge about humans as social beings to improve computational analysis in their ground-breaking work on bounded rationality, Soar, and the book "Unified Theories of Cognition". Today, such sophisticated cognitive models are used in applied settings such as the team soar models. Similar developments have occurred in the other area. Thus there has been a movement from studies of interaction and task to tools for team design, and from studies of chance to studies of complexity.
Computational organizational models are often developed and evaluated by multi-disciplinary teams work synergistically making advances in both social science and computer science. Scientists conduct virtual experiments within computational laboratories to examine the effect of fundamental social processes on group behavior. This research makes contributions to mainstream social science, management science, artificial intelligence and computer science. This is done by fostering progress on such issues as: fundamental principles of organizing, large scale qualitative simulation, comparison and extension of constraint based optimization for extremely complex and dynamically changing performance surfaces; optimal and flexible coordination structures for different types of agents (humans, corporations, WebBots, or robots) and tasks aggregation and disaggregation of distributed objects; coordination algorithms; virtual and web-based testbeds for determining the social impacts of information technology and its prospects for diffusion, organizational and multiagent learning; the evolution of groups, networks, and markets, the tradeoff between agent quantity and computational complexity, he development of computational tools for assessing knowledge networks, communication networks, team mental models, and group effectiveness, the development of easy-to-use cost-effective computational tool kits for designing and building agent based models of organizations, teams, markets, and social systems.
In this seminar, the following issues are discussed: 1) The nature of
computational organizational models; 2) A set of classic models and the core
findings; 3) Virtual experiments; and 4) A brief overview of validation.
