[Mortality analysis in studies of industrial cohorts. III. Testing the trends of standardized relative risks]

The present paper covers the third part of a larger work concerning mortality measures applied in the studies on industrial cohorts. The aim of this paper is to discuss the problems of testing the trends of standardized relative risks (SRR). In spite of its advantages the standardized relative risk (SRR) is not often used in empirical works to assess mortality. If it is used at all, it is usually based on the chi-square test. The authors of this paper must state that, as it results from their own experience, a non-parametric test is more useful in evaluating mortality than chi-square or F test.