A Tukey type trend test for repeated carcinogenicity bioassays, motivated by multiple glyphosate studies
In the last two decades, significant methodological progress to the simultaneous inference of simple and complex randomized designs, particularly proportions as endpoints, occurred. This includes: i) the new Tukey trend test approach, ii) multiple contrast tests for binomial proportions, iii) multiple contrast tests for poly-k estimates, and Add-1 approximation for one-sided inference. This report focus on a new Tukey type trend test to evaluate repeated long-term carcinogenicity bioassays which was motivated by multiple glyphosate studies. Notice, it is not the aim here to contribute to the evaluation of Glyphosate and its controversies. By means of the CRAN-packages tukeytrend, MCPAN, multcomp the real data analysis is straightforward possible.
READ FULL TEXT