Social Simulation using R

Lyndon Walker
Department of Statistics and Department of Sociology, University of Auckland
lk.walker@auckland.ac.nz


Social simulation provides an effective means of observing the effects
of complex social interactions and decision making which are too
complex to be modelled by traditional statistical techniques.  It is a
technique which is growing in popularity, however, it is more
regularly utilised by social scientists than statisticians.  As a
result, the software which has been developed for social simulation
tends to lack some of the flexibility and analytical power that we see
in statistical programmes like R.  This paper examines the conversion
of a social simulation model from a common freeware simulation
programme into R.  As a programming tool, R provided a simple
interface and a familiar and flexible language which made the creation
of a social simulation programme as easy as it was with the dedicated
simulation tool, but it also provided a plethora of useful statistical
functions with which to analyse the simulation output.  By
demonstrating the ease and effectiveness with which social simulation
can be performed in R, I hope to be able to promote R to a broader
audience of social scientists who would not normally use it.