Professor Daniel Farewell

Overview

I am interested in the development of novel statistical theory and methodology, especially for causal inference and missing data. I collaborate with colleagues in Cardiff and beyond in applications to biomedical science, often using anonymised record-linked databases.

I maintain a particular focus on the analysis of longitudinal data, where study subjects may be observed on multiple occasions. Such analyses are especially challenging when the observed timings of measurements can provide further information about study subjects. For instance, if (when compared to healthier participants) those subjects who are most unwell tend to be observed less frequently, or to drop out earlier, careful analysis may be needed to avoid overrepresenting the healthy subjects when drawing conclusions from the study.