Values (x) are sorted from smallest to largest. Then Z values are calculated as follows:
Zn-i+1=[i*(n-i)(Xn-i+1 - Xn-i)]^0.5
where n is the sample size
for i = 2,...,n calulate the 25 percent trimmed mean and divide into Z.
This standardizes the distribution of the weighted gaps around a middle value of one. Suspiciously large
observations should correspond to large standardized weighted gaps.