Abstract
When studying the effect of a factor of risk on the incidence of a critical event, the commonly used 2x2 exposure-affection table reduces all the sample information to the pertinence of each individual to one of four table cells (A= exposed and affected, B= exposed not affected, C= affected not exposed, D= neither exposed nor affected). Therefore, we have to define two alternative levels of exposure and affection, usually divided by a threshold value. For instance, a person may be classified as “moderate smoker” if he smokes less than 5 cigarettes a day. Nevertheless, the degree of exposure may be sensibly different between individuals, and also the level of disease (or critical event) does often show a relevant variability between units. Recently, one of the Authors (Brizzi, 2007) has proposed a fuzzy pattern in order to define a rescaled 2x2 contingency table, taking into account the presence of a partial level of exposure and/or affection. Here, such a pattern is briefly outlined and applied to a set of data, related to perceived health status of aged people living in the suburban area of Bologna