Risk factors have not been completely accepted in public health and clinical medicine, largely because of ambivalence about inferential statistics. Their prominent role in the health education movement has raised several fundamental issues, including individual versus social responsibility for disease and population-wide- versus high-risk strategies.

Risk factors have brought public health and clinical medicine closer together than ever before. The role of public health is to identify risk factors, educate the public about prevention and treatment, and promote changes in individuals and public and private organizations. The role of clinical medicine is to diagnose risk factors in individual patients and treat them through lifestyle changes or medications.

The identification and treatment of risk factors rely on inferential statistics, which has become the standard method of finding relationships in samples of persons and generalizing the findings to populations of interest. Inferential statistics has had a long record of achievements in the social sciences, agriculture, education, and business, where it produced a fundamental reorientation of their conceptual frameworks. Probabilistic modes of analysis replaced deterministic ones, multifactorial etiology replaced the analysis of one causal factor at a time, and correlations replaced cause-andeffect relationships as the objectives of research investigations.

In public health and medicine, inferential statistics has had to compete with older methodologies: clinical observation, pathology, laboratory investigation, and vital and other descriptive statistics. These methodologies produced revolutionary and profound discoveries in disease etiology, prognosis, and therapy. Their successes have made them the criteria by which all new methodologies are judged. They dominate the philosophical and methodological bases of professional education; they are pervasive in journals and conferences; they serve as the evidence of last resort in controversies.

Inferential statistics was brought into public health and medicine partly because of the limitations of the dominant methodologies when applied to chronic and degenerative diseases. Clinical observation, pathology, and laboratory investigation could not explicate the etiology of diseases with extremely long latency periods.