Causal inference in epidemiological research

Traditionally, statistics has been viewed as the branch of science which deals with association. Many epidemiological research questions, however, are concerned with causation, not association. In this thesis we develop novel statistical methodology to address four epidemiological problems properly, from a causal inference point of view. We show, that for these four problems, our methods offer an attractive alternative to the `standard' methodology, which may not yield the desired (causal) inference.

List of scientific papers

I. Sjölander A, Humphreys K, Palmgren J (2008). On informative detection bias in screening studies. Stat Med. 27(14): 2635-50
https://pubmed.ncbi.nlm.nih.gov/17918781

II. Sjölander A, Humphreys K, Vansteelandt S, Bellocco R, Palmgren J (2008). Sensitivity Analysis for Principal Stratum Direct Effects, with an Application to a Study of Physical Activity and Coronary Heart Disease. Biometrics. Aug 27: Epub ahead of print
https://pubmed.ncbi.nlm.nih.gov/18759834

III. Sjölander A (2008). Bounds on natural direct effects in the presence of confounded intermediate variables. Stat Med. Nov 26: Epub ahead of print
https://pubmed.ncbi.nlm.nih.gov/19035530

IV. Sjölander A, Humphreys K, Vansteelandt S (2009). A principal stratification approach to assess the differences in prognosis between cancers caused by hormone replacement therapy and by other factors. [Submitted]