Pooled inference across sexes for the in vivo micronucleus assay.

The Japanese Environmental Mutagen Society has investigated the extent of sex differences in the in vivo micronucleus assay (Sutou et al., 1986). In light of their findings, this manuscript reexamines the statistical analysis of the assay. A test statistic which pools the inference over sexes is introduced. The sensitivity of this statistic is examined in comparison with the more traditional procedure of analysis within each sex. The impact of extra-Poisson variation among animals on the validity and sensitivity of the test procedures is also examined.