Do fire trucks cause fires? Counterfactuals, Statistical models, and Causal inference from observational data.

Albyn Jones

Counterfactual formulations of causality can be traced as far back as Hume. Counterfactual statistical models were first proposed by Jerzy Neyman in a recently rediscovered paper of 1929. Don Rubin and others have popularized the models and stressed their importance in the analysis of causal inferences from observational data. This talk will be a non-technical survey of the problems of causal inference from observational data, and the practical implications of the counterfactual models for inference.