Cultured cells homozygous for a mutation show a significantly different phenotype from cells homozygous for the wild type allele. Both have normal standard deviations. However, when I measure the same phenotype in cells heterozygous for the mutation, the mean is as expected between both measurements, but the standard deviation increases to +/- 40-50%. What could be the reason? Is it possible that the cells express different ratios of both alleles although they have the same genetic background?