On Nei and Roychoudhury's sampling variances of heterozygosity and genetic distance.

NEI and ROYCHOUNHARY have developed an inequality relationship between the expected values of (j-x - g)-2 and (ĝ - g)-2 where j-x is a biased and g is an unbiased estimate of population homozygosity g. Their inequality is somewhat weak and can be improved, as is demonstrated in this paper.