Social dynamics of substance use through minds and models

Progress in psychological theory has slowed in recent decades, while advances in computational power have opened new avenues for studying systems that represent human behaviour. In this thesis, I focus on the formalisation of psychological theories, which has several advantages: mathematical models require clear definitions, helping to clarify ambiguities in constructs and theories. These models can also generate precise predictions for hypothesis testing and guide future empirical research. In addition, formal models can simulate complex systems and show how small interactions or changes can lead to emergent behaviour. Thus, by studying these models, we can gain a deeper understanding of human behaviour and develop more effective interventions.
In particular, I looked at substance use its interactions with ones social environment. A review of recent formal models in these areas highlighted a lack of integration between psychological and social perspectives. To address this, I developed an epidemiological model that treats both heavy drinking and abstinence as diseases that spread within a social network. This highly abstracted model still accurately predicts empirical population-level temporal data. Furthermore, I applied innovative psychometric methods to this large dataset to explore correlations between drinking and various predictors of substance use, including smoking, occupational status and social environment. Finally, I collaborated on a more extensive formal model of substance use that integrates and clearly defines common psychological constructs, effectively bridging contemporary psychological theory and social interactions while predicting individual-level outcomes.