Making Inference and Making Decisions
The designs described in Chapter 2 allow you to estimate associations between the exposures under study and the corresponding health outcome. Making causal inference is a more complicated issue. No design and no statistical procedures will in themselves allow you to make causal inference. You can only be more or less certain about causal relations, and making causal inference rests to some extent on subjective consideration. However, the opinion of those who know the subject matter well carries more weight than the opinion of ignorants, although the latter could be right. History shows that the skillful more often got it right, because knowledge allows you to identify some non-causal associations. Skills will allow you to put hypotheses to critical tests that often will reveal associations caused by bias or confounding.
KeywordsCausal Inference Coffee Consumption Inadequate Evidence Etiologic Fraction Subjective Consideration
- 3.WHO, IARC. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 78, Ionizing Radiation, Part 2: Some Internally Deposited Radionuclides. IARC Press, Lyon, France, 2001.Google Scholar