Studying pathways between social and biological factors: can modern methods in causal inference help?
When: Tuesday 11 September 2012
At: 56th Annual Scientific Meeting of the Society for Social Medicine, 2012
Where: London School of Hygiene & Tropical Medicine, Keppel Street, London, WC1E 7HT
Twitter: @pathwaysNCRM
Summary:
To study the influence of socio-economic and socio-demographic factors on health outcomes we must recognise the additional potential biases that are encountered when including factors lying on presumed pathways in regression models. It may be that more sophisticated analytical approaches than those traditionally employed in epidemiology and social science are necessary.
In this meeting we will use concepts, tools and methods arising from the causal inference literature (such as counterfactuals, causal diagrams, and the g-formula) to gain insights into the underlying assumptions of traditional methods used to study mediation, and explore more modern methods in which some of these can be relaxed.
