Simulating dynamic systems in health psychology : a thesis presented in partial fulfilment of the requirements for the degree of Master of Arts in Psychology at Massey University, Palmerston North, New Zealand

Abstract

Despite their advocacy of the biopsychosocial model, health psychologists use a relatively narrow repertoire of techniques for developing and testing theory. These techniques have limited application to research questions concerning phenomena that are multidimensional, multilevel and change over time. This thesis demonstrates an alternative, dynamic systems approach to such questions in health psychology. It introduces some ideas in systems and dynamics and how we might model these. It uses an example to demonstrate the use of these ideas to develop a dynamic systems model in a health psychology context. The example is drawn from the epidemiological finding of a positive correlation between income inequality and mortality, and the proposal that this relationship may be mediated by processes that result in social disruption. The thesis explores the construction of a dynamic systems model to examine how a change in income inequality might affect the network of social relationships in a population. Social relationship processes in the model are based on some findings from social psychology, and these are incorporated into a network model, which is realised as a computer simulation. Simulation runs suggested that an increase in income inequality can produce a ripple of relationship breakdowns. Contrary to intuition, the number of relationships lost was limited if the change was introduced suddenly, and if there was a high rate of making and breaking relationships. Further, reversing the change did not reverse the loss of relationships. The development process and the results obtained are discussed, and it is argued that dynamic systems simulation may be useful for developing and testing theory that applies to multilevel, multidimensional processes in health psychology.