Popis: |
For the BVE (BiVariate Exponential) distribution, defined and explained by Marshall and Olkin (1967), the population Kendall's τ (1948), which is the probability of concordance minus the probability of discordance for a bivariate distribution, is shown to be equal to the population correlation coefficient. Moreover, Kendall's t statistic is superior to Pearson's r for estimating the BVE correlation. Some other bivariate distributions, due to Gumbel, Morgenstern and Freund, are also considered. |