Epidemiologie et statistique medicale: elements de methodologie [1]

ABSTRACT

Because of the variability of the studied material, the biomedical research requires an approach based upon the average properties of a group rather than the individual Reponses. The statistics are the very method that answer this requirement. However, the statistical interpretation can be valuable only if the research has been correctly conceived, and particulary if the definition of the problem, the hypotheses and the study protocol have been clearly established. This protocol will define the sort of study adapted to the problem i.e. a descriptive study [cross-sectional unliked studies], etiological [follow-back case-control or prospective studies] or'an evaluating research according to the cases