Confounding and Bias
In Part II we described how to estimate a measure of association between an exposure and an end point, often a disease. We are interested in “effects” but are only able to measure associations, and we have several procedures to go through before associations can be called “effects.” Measures of association emerge for all possible reasons and only some of these relate to “effects.” Effect is a causal word to be used with care.
KeywordsLung Cancer Risk Inductive Inference Mycosis Fungoides Healthy Work Effect Descriptive Epidemiology
- 1.Last JM (ed.). A Dictionary of Epidemiology, 3rd Edition. Oxford University Press, New York, 1995.Google Scholar