A method to apply the robust estimator of dispersion, Qn, to fully-nested designs in the analysis of variance of microbiological count data.