ABSTRACT
In this chapter, I will look at some examples from the history of statistics— examples which help to define problems of causal inference from non-experimental data. By comparing the successes with the failures, we may learn something about the causes of both; this is a primitive study design, but one that has provided useful clues to many investigators since Mill (1843). I will discuss the classical research of Pierre Louis (1835) on pneumonia, and summarize the work of John Snow (1855) on cholera. Modern epidemiology has come to rely more heavily on statistical models, which seem to have spread from the physical to the social sciences and then to epidemiology. The modeling approach was quite successful in the physical sciences, but has been less so in the other domains, for reasons that will be suggested in Sections 4–7.
