1. In Freedman’s 1991 article, “Statistical Models and Shoe Leather”, Freed- man put forth as an example of exemplary study, John Snow’s investi- gation of the causes of Cholera. The following questions discuss Snow’s investigation into the causes of Cholera.
(a) John Snow essentially uses a comparison of water provided by the Lambeth Water Company and the Southwark and Vauxhall com- pany as a natural experiment to investigate the causal impact of dirty water as a conduit for Cholera. First, if you could devise a random- ized controlled trial to test this hypothesis? How would you do so? What issues would randomization potentially surmount? How would compliance with assignment to treatment possibly contaminate your results?
(b) Consider Table 1 (Snow’s Table IX) on page 298 of the manuscript. Explain how the data presented in the graph is consistent with the hypothesis is consistent with the hypothesis that dirty water is the conduit via which Cholera is transmitted?
(c) What are some issues that Snow identified that possibly confound the hypothesis that dirty water is the conduit via which Cholera is transmitted?
(d) What evidence does Snow offer to support the assumption that, when examining Lambeth and Southwark and Vauxhall water distribution, the potential confounders are not an issue?