Workshop George Davey Smith - Triangulation of evidence and causal inference in epidemiology

Workshop

“This workshop will cover issues relevant to applied causal inference in epidemiology, building on practical examples across diverse fields. The value of formal approaches to quantitative causal inference will be explored in the context of such practical examples. The use of falsification methods (including negative control exposures & outcomes and cross-context comparisons) and the harnessing of external perturbations (including formal instrumental variables analyses) will be discussed. A triangulation approach based on a pre-specified set of methods – each which may be biased, but with biases anticipated to be near-orthogonal – will be advanced. The notion that causal inference will always be a joint qualitative and quantitative activity, and that quantification of causal estimates should go as far as is useful, but no further, will be defended. The focus will be mainly on non-genetic approaches to causal inference, as a second workshop will cover Mendelian randomization in detail.”

For questions, please contact Lisette Kuil (l.kuil umcg.nl)

This workshop is part of a series by George Davey Smith:
14 october: Workshop 'Mendelian Randomization'
16 october: Triangulation of evidence and causal inference in epidemiology
17 October: Seminar 'Post-Modern Epidemiology: When methods meet matter'