Statistical significance multiplicity

When the data as a whole are reviewed for studies on humans exposed to ethylene oxide, no conclusion can be made that there is an increase in mortahty associated with those exposed to ethylene oxide. Two Swedish studies (247,248) indicated an increase in leukemia for workers exposed to multiple chemicals including ethylene oxide however, in a recent larger Swedish study (249) of workers exposed to only ethylene oxide, there was no association of any type of cancer increase for these workers. In a recent study sponsored by NIOSH, there was no significant increase in mortahty observed for cancer when all types are combined or for certain individual types of cancer, even for those people who worked the longest and were observed the longest. However, a statistically significant increase in mortahty from certain types of lymphoma was observed for male workers. This is contrary to the results observed for female workers. In addition, four other cohort studies of ethylene oxide-exposed workers have been pubhshed (250—253), but no unequivocal increase in the risk of cancer was observed. 

The analysis of rank data, what is generally called nonparametric statistical analysis, is an exact parallel of the more traditional (and familiar) parametric methods. There are methods for the single comparison case (just as Student s t-test is used) and for the multiple comparison case (just as analysis of variance is used) with appropriate post hoc tests for exact identification of the significance with a set of groups. Four tests are presented for evaluating statistical significance in rank data the Wilcoxon Rank Sum Test, distribution-free multiple comparisons, Mann-Whitney U Test, and the Kruskall-Wallis nonparametric analysis of variance. For each of these tests, tables of distribution values for the evaluations of results can be found in any of a number of reference volumes (Gad, 1998). 

Bruntz et al. applied multiple regression analysis and found that the method of least squares yielded a set of coefficients that produced a 0.84 correlation of ozone concentration with the data. Adding mixing height to the correlation yielded no statistically significant improvement in agreement with the assertions of Hanna. ... 

---