A comparison of the average degree of nonrandomness of patchiness in two or more populations as well as the degree of their statistical accuracy is often desired. Both Jackknife and Bootstrap provide nonparametric estimation of standard error of an estimator and thereby making possible the significance tests and the establishment of confidence intervals and testing the significance. In this paper we examine the statistical accuracy (standard error, bias, mean squared error) and provide methods for estimating confidence intervals for three indices of spatial dispersion pattern.
