What is Simulation Modelling

Computational simulation is a methodology for creating and executing models. Initially, simulations have been developed and used only in the realm of "hard" sciences (physics, chemistry, biology). Social sciences, where observations in laboratories are very expensive, if not impossible, are beginning to exploit them too. In particular, organizational studies tend to use agent-based simulations, software that simulates the behavior of individual "agents" by setting their attributes and some rules of interaction. Behaviors and their evolution emerge from several iterations of the model embedded in the software. This permits both to validate existing theories and to create new ones.
Compared to traditional research methodologies social simulation ensures several advantages:

1. Simulations force theorists to create rigorous models: in a simulation, a model must be constructed with explicit hypoyheses and clear transition rules.

2. Simulations allow theory testing: on one side, simulations can incorporate real data inside formalised models for empirical research can provide realistic parameters. On the other side, simulations can address the fieldwork towards new directions of research.

3. Simulations can be used to develop social experiments: by changing control variables, simulations can be used to re-create manipulations used in experimental methodologies.

4. Simulations permit the analysis of the evolution of the variables under examination: this has the evident advantage of saving the high costs of longitudinal analyses of empirical research.


Knownetlab and organizational simulations

Organizational simulations are still very innovative tools for understanding organizations. They are critical to Knownetlab as can be applied in all the research/consultancy themes of our group: design of processes and structures, impact of information and communication technologies, knowledge management and territorial marketing.

Knownetlab applies the language LSD (Laboratory on Social Simulation) for its social simulations. LSD has been developed at University of L'Aquila. This C++ based language provides a simple editing interface for the development of functions and models. The advantage of LSD is the high operational flexibility (as in more complex languages like SWARM) and the concurrent user-friendliness (as in simpler software like Blanche and SimVision).

For a basic list of references, see:

- L. Biggiero, 2009. A short presentation of three (relatively new) methodologies: multicriteria decision making; social network analysis, and agent-based simulation modelling. New methods MCDM SNA ABSM [slides]
- Biggiero L. 2009. Basic issues and examples in the agent-based simulation modelling. [slides]
- Carley, K.M. and Prietula M.J. (eds.) 1994. Computational organization theory. Hillsdale (NY): Lawrence Erlbaum.
- Epstein, J.M. 2005. Generative social science: studies in agent-based computational modeling. Princeton: Princeton UP.
- Epstein, J.M. and Axtell, R. 1996. Growing artificial societies: social science from the bottom-up. Cambridge (MA): MIT Press.Gilbert, N. 2008. Agent-based  models. London: Sage.
- Gilbert, N. and Troitzsch K.G. 2005. Simulation for the social scientist (second edition). NY: Open University Press.
- Lomi, A. e Larsen, E.R. 2001. Introduction, in Lomi A. and Larsen E.R. (eds.) Dynamics of organizations: computational modeling and organization theories. Cambridge MA: MIT Press, pp. 3-36.
- Tesfatsion, L. and Judd K.L. (Eds.) 2006. Handbook of computational economics (vol. 2). Amsterdam: North Holland.

Models