Snow, in contrast to Farr's epidemiology, was far more innovative and spontaneous in his methods, which also made his conclusions, in the eyes of his colleagues more suspect. As well as doing his own hands-on research, Snow analyzed the "natural experiment created when one water- supply company of London, the Lambeth Company -- but not the Southwark and Vauxhall Company -- moved its water inlet to a less polluted area of the Thames. Snow's hypothesis was that if cholera was related to consumption of water contaminated by human excrements, then mortality rates should be greater among those who drank the contaminated water supplied by the Southwark and Vauxhall Company than among those who drank the cleaner water supplied by the Lambeth Company" (Morabia 2001: 224). Determining the exact purity of the water supply at a given point in time, however, was difficult, and made it difficult for Snow's thesis to eliminate other possibilities of causation (Eyler 2001: 227-288).

Thus from the point-of-view of his colleagues at the time, there was some 'sloppiness' to Snow's published methods and his unequivocal claims, in marked contrast to the cleaner, more balanced portrayal provided by Farr. "Snow did not know the number of people at risk of cholera in his test case…he did not even know the number of households supplied by each company in the districts with the mixed water supply" (Eyler 2001: 227). Because the experiment was based upon real-world data, Snow's experimental analysis was not perfectly controlled. There were distinct demographic differences thought to be relevant at the time between the populations using the different water supplies (Eyler 2001: 227). Snow also did not entertain a multifactoral possibility for the spread of disease. On the surface, Farr seemed more balanced and scientifically impartial: Snow was more passionate and less rigorous in his defense of his belief schema.

In defense of Snow's critics, one of the most common criticisms of many scientific experiments is the unwillingness on the part of the researchers to deal with contradictory data that does not support their original hypothesis. Farr's tentativeness made him seem more fair and scientific. Yet while Snow was criticized for subsuming too much evidence to suit his thesis, his critics could and should be equally criticized for subsuming their view of his research to suit their own ideas about how all disease were airborne.

Even if Snow may have been more dogmatic in his presentation,...

For Snow, the purpose was to verify his thesis regarding a public health crisis, not merely the gentlemanly pursuit of discovery (Eyler 2001: 230). Eventually, after more reasoned consideration Farr began to change his mind and concede that waterborne transmission of the ailment was possible (Eyler 2001: 230). But without Snow's dogmatism, it is unlikely that such a view would have gained in popularity -- just as it is true that without Farr's statistics, Snow may not have come up with his thesis. The Snow vs. Farr debate suggests that the world needs both Farrs and Snows -- data-based researchers and as well as passionate, hands-on testers of theories. Both are needed to get to the bottom of disease causation, given the challenges that causality poses to scientific investigators of disease-related phenomena.
References

Dallal, Gerard E. (2008, September 12). Cause and effect. The little handbook of statistical practice. Retrieved August 30, 2010 at http://www.JerryDallal.com/LHSP/cause.htm.

Eyler, J.M. (2001). The changing assessments of John Snow's and William Farr's cholera studies. Soz Praventiv Med, 46 (4): 225-232. Retrieved August 30, 2010 at http://www.epidemiology.ch/history/papers/eyler-paper-1.pdf.

Morabia a. (2001). Snow and Farr: A scientific duet. Soz Praventiv Medicine, 46 (4): 223-224.

Retrieved August 30, 2010 at http://www.epidemiology.ch/history/papers/SPM%2046(4)%20223-4%20Morabia%20Editorial-2.pdf.

Schoenbach, Victor. (1999). Causal inference. EPID. 168, 285.

Snow, S. (2004) Cholera, chloroform, and the science of medicine: A life of John Snow.

The New England Journal of Medicine, 350(1):90. Retrieved August 30, 2010

from http://proquest.umi.com/pqdweb?index=4&did=522380981&SrchMode=1&sid=2&Fmt=4&VInst=PROD&VType=PQD&RQT=309&VName=PQD&TS=1266797764&clientId=29440

Wilson, E. Bright. (1952). An introduction to scientific research. New York: McGraw-Hill.

UCLA Department of Epidemiology School of Public Health (2001). William Farr:

campaigning statistician. Retrieved August 30, 2010 at from http://www.ph.ucla.edu/epi/snow/farr/farr_compiler.html.