Inequalities in physical activity as an emergent feature of a complex system: an agent-based modelling approach

The agent-based model was developed in four main stages: (1) conceptual model development, (2) operationalisation, (3) parameterisation and calibration and (4) consistency analysis. The conceptual model demonstrates how multiple levels of income, psychosocial, physical, and perceived environment factors influence an individual’s decision to practice leisure-time physical activity. Incorporating this conceptual model to directly inform the agent-based model, we were then able to produce inequalities in leisure-time physical activity, consistent with patterns reported in the literature.

Through a series of scenarios designed to optimise interventions, the model provided a number of key findings which can be useful insights for intervention design and optimisation. In conclusion, this thesis demonstrates that inequalities in leisure-time physical activity can be interpreted as an emergent feature of a complex system. Through the conceptual model development, along with the subsequent agent-based model development and analysis, we demonstrate the importance of appropriately considering intervention design, implementation, and appropriate evaluation strategies to allow for more effective, feasible and maintainable ways to promote physical activity and reduce inequalities in LTPA.<br><br>