Abstract
In econometrics it is important to distinguish whether two variables are correlated or not. Distribution of correlation coefficient of n pairs of samples from two independent normal populations is theoretically known. Based on the distribution and the value of the sample correlation coefficient we decide whether the two variables are correlated or not. We compute distribution of correlation coefficient of n pairs of samples from two independent uniform populations by computer simulation. We see that we cannot distinguish the two distributions of the correlation coefficients when n<11. When n>11, however, the two distributions are different definitely.
