Selection Bias, Missing Data and Causal Inference

Causal inference can be attempted using different statistical methods, each of which require some (untestable) assumptions. Common methods include multivariable regression, propensity scores, g-methods (no unmeasured confounding) and instrumental variables (no association between instrument and outcome, other than via the exposure). Less attention has been given to the impact of selection (e.g. selection into a study, analysis of cases only) or missing data (e.g. dropout from a study, death due to other causes) on different methods for causal inference. Using directed acyclic graphs (DAGs) I will discuss some of the ways in which bias can occur due to selection or missing data, and methods that might be used to detect or mitigate against this bias. Applied work shows evidence of non-random selection into and dropout from studies including ALSPAC and UK Biobank, and I will discuss how this might impact causal analyses using these datasets.

Please join the event.

About Kate Tilling

Kate Tilling is Professor of Medical Statistics at the University of Bristol and an MRC Investigator. Following a degree in Maths, MSc in Applied Statistics and PhD in Epidemiology she took up a post as lecturer in Medical Statistics at King’s College London, moving to the University of Bristol in 2002. She has subsequently built an interdisciplinary research team in the MRC Integrative Epidemiology Unit and leads a MRC-funded research programme on the development and application of statistical methods for causal inference.