Experimental Simulation of a Social System

Abstract

Operational research has had only limited success in tackling problems which involve dynamic social factors. The social sciences today are not oriented to contribute to the solution of these problems. Rather than abandon the problems, or deal with them inadequately by ignoring the social factors, there is a third possibility. The situation can be simulated, with necessarily simplified rules of social action and interaction. The simulation should be regarded as a tool for experimentation rather than as an exact description, and progressive sophistications of the representations of the social factors can be developed. Illustrations of the adequacy of this approach are drawn from a model of the prescribing habits of doctors.

^†This work was carried out when the author was employed by Science in General Management Ltd.

^†This work was carried out when the author was employed by Science in General Management Ltd.