Epistemological Aspects of Simulation Models for Decision Support

Epistemological Aspects of Simulation Models for Decision Support

Anthony H. Dekker (Graduate School of Information Technology & Mathematical Sciences, University of Ballarat, Ballarat, VIC, Australia)
Copyright: © 2013 |Pages: 23
DOI: 10.4018/jats.2013040103
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

In this paper, the author explores epistemological aspects of simulation with a particular focus on using simulations to provide recommendations to managers and other decision-makers. The author presents formal definitions of knowledge (as justified true belief) and of simulation. The author shows that a simple model, the Kuramoto model of coupled-oscillators, satisfies the simulation definition (and therefore generates knowledge) through a justified mapping from the real world. The author argues that, for more complex models, such a justified mapping requires three techniques: using an appropriate and justified theoretical construct; using appropriate and justified values for model parameters; and testing or other verification processes to ensure that the mapping is correctly defined. The author illustrates these three techniques with experiments and models from the literature, including the Long House Valley model of Axtell et al., the SAFTE model of sleep, and the Segregation model of Wilensky.
Article Preview

1. Introduction

In recent times, there has been a growing interest in epistemological aspects of simulation (see e.g. Becker et al., 2005). Do simulation produce knowledge, and if so, how? Some writers suggest that simulations raise new epistemological issues, while Frigg and Reiss (2009) argue that “simulations ... raise few if any new philosophical problems. The philosophical problems that do come up in connection with simulations are not specific to simulations.” In other words, epistemological analysis of simulations can draw on studies of the philosophy of knowledge going back to Plato’s Theaetetus. In this paper, we explore epistemological aspects of simulation with a particular focus on simulations which are used to provide recommendations to managers and other decision-makers, especially in the area of organizational and social modeling. Should these decision-makers ask the question “Do you really know that?” then the simulation modelers must have a reply to hand.

Under what circumstances do simulation models produce knowledge? Taking the classical definition of knowledge as justified true belief, we outline some basic epistemological principles, and describe a number of techniques by which simulations can provide knowledge to decision-makers. In particular, we consider how the architecture or theoretical basis of a model may be justified, how the parameter values “plugged into” a model may be justified, and how verification can justify the belief that a model has been implemented correctly. We illustrate these techniques with some simple simulations, including the Kuramoto model of coupled oscillators (Strogatz, 2000), the Long House Valley model of Axtell et al. (2002), and the Segregation model of Wilensky (1997b). First, however, we must clarify precisely what we mean by “knowledge.”

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 9: 2 Issues (2017): Forthcoming, Available for Pre-Order
Volume 8: 1 Issue (2016): Forthcoming, Available for Pre-Order
Volume 7: 3 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing