Answers are grounded in this paper and cite their sources.
The paper introduces a x2-approximation to multivariate kurtosis b2,punder normality. It requires calculating the third moment of b2,pwhich is obtained. We compare the approximation with simulated percentage points and the normal approximation, and find it to be adequate for p=l and 2. For p=3, the simple average of this estimate and the normal approximation is found to be generally superior to either approximation on its own. For p=4, the normal approximation is best for non-extreme values of ∝
No takes yet. Share an insight, caveat, or question.
Mardia et al. (1983) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: