Skewness and Kurtosis


New Member
I have to 'show the standard error of the skewness and kurtosis as a footnote to a table' - any ideas on how this is expressed? I thought it might be quoting the number of standard errors, but apparently not! Any advice on how people normally show this would be great :)

I have come across this problem once before, and in doing some research, I could only find an approximated standard error for each. For the skewness:

SES = sqrt(6/n)

where n is the sample size. As a general rule, this approximation is best when n > 150. For the kurtosis:

SEK = sqrt(24/n)

and the general rule here is n > 1000

Hope this helps.