Calculate skewness and kurtosis

General
#1

Hello,

I get different results for calculating skewness and kurtosis when using stat.desc (pastecs package) and norm (QuantPsyc Package). Does anybody know why? And which calculation is correct?
I attached a screenshot of the results with both methods for my variable "UW".

Kind regards
Christiane

Bildschirmfoto 2021-02-14 um 17.44.46
Bildschirmfoto 2021-02-14 um 17.44.461053×240 26.8 KB

#2

Welcome to the community!

I think you should use stat.desc if you are looking for descriptive statistics. The norm is testing whether the distribution is symmetric or not, and I don't think it uses \gamma_1 itself as the statistic.

Take a look at these source codes:


1 Like