Föreläsningar och seminarier CAUSALab Methods Series at IMM
Effects of smoking on behaviour and mental health: A case study in the need for triangulation
Speaker: Marcus Munafò, University of Bristol
Abstract: Smoking is strongly associated with a number of adverse mental health outcomes, but whether this reflects a causal pathway (and, if so, in what direction) remains an open question. Various techniques allow for stronger causal inference when using observational data. These include instrumental variable methods (in particular Mendelian randomisation, which uses genetic proxies for an exposure of interest as instruments), but also cross-contextual comparisons and the use of positive and negative controls. No single approach will always provide the correct answer; therefore, robust conclusions will only emerge where there is converging evidence from multiple approaches that have different strengths, weaknesses, and sources of bias. I will discuss evidence that smoking plays a causal role in poor mental health in the context of this triangulation framework.
Sign up: https://survey.ki.se/methodsseries