Update: An interactive version of the central model of my thesis is now here
Small Worlds, Sour Grapes and Social Deprivation: the Computational Application of Peter Hedström’s DBO Theory
In Dissecting the Social, Peter Hedström proposes an agent-based sociological theory in which each individual’s action is determined by their desires, beliefs and opportunities. He then uses this model to simulate social interaction, and reaches certain conclusions about the macro-social consequences of different forms of agency. While the DBO theory itself is psychologically compelling, the parameters of the model that are used to apply it are very restricted. By varying the assumptions made in the practical application of the theory, I hope to gain more insight into the predictive possibilities and theoretical implications of this agent-based model.
I wrote this undergraduate thesis as part of my BA in Philosophy, Politics and Economics at the University of Oxford, under the supervision of Peter Hedström.
The main parameters that I varied were the nature of individual decision-making mechanisms and the types of networks on which agents interact.
The simulations were run using a number of Python scripts using the python-graph and pydot libraries, outputting to GraphViz and GnuPlot. For what it’s worth, the scripts written for these simulations come to 4,284 Source Lines Of Code (a statistic generated using David A. Wheeler’s ‘SLOCCount’.)
The thesis won the Gibbs Thesis prize in Politics from the University of Oxford.
The thesis is not yet available for download, but may be soon.