Hmmm, I can think of quite a few distributions that have a median/mean ratio of exactly 1 and are distinctly not normal. I can't imagine making a formal test of normality using such a ratio.
Isn't the range of a normal distribution rather large? The tests of normality (which have, IMHO, little if any utility) with which I am familiar compare the sample cumulative distribution with that which would be expected were the sample drawn from a normal distribution -- I've seen this done with Chi-square and Kolmogorov-Smirnov, for example. Karl W. ----- Original Message ----- rule of thumb for testing normality is to divide the median by the mean, and if the ratio is between 0.9 and 1.1, the distribution is approximately normally distributed. a test for normality based on the range divided by the standard deviation Does anyone know of a formal test for normality employing the mean and median, i.e. the central tendency instead of the dispersion? . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
