Every Statistical Problem is a Missing Data/Causal Inference Problem

Across numerous fields of domain science research, challenges arise routinely where data of interest are missing and/or inferences are required regarding, for example, the causal effects of treatments based on observational data sources. Accordingly, there is an enormous literature on statistical methodology for analysis in the presence of missing data and for causal inference. It is in fact the case that many statistical problems can be thought of as missing data/causal inference challenges, which can suggest fruitful ways to develop methodology to address them. I will discuss several projects in which casting the statistical problem as one of missing data/causal inference has led to new methodology.

Share this Event