hello every one. i have this output abt skewness and kurtosis, i confused bcs some articles said that we just look at value of skw and kr, but other said that shoul to see by z value (value skw or kr divided by st error). what ois the right method to see the normality using skw and kr? (btw my samples are 381)
You probably don't want a z-test for skew or kurtosis. You have the estimates of the skewness and kurtosis values. Doesn't that tell you what you need to know ?
There are cases where you may want a test for these, but, in your case, do you ?
The two methods may be applicable and almost give the same interpretation.
In your case skewness is less than .5 ssuggests the data is nearly symmetric.
When calculate z = .1 /.125 = .8 which is less than 1.96 for 95% c.l. so also this suggests symmetry.
While keratosis is more than 3 with negative charge suggesting platykurtic distribution. When calculate z= -.52-3/.249 = - 14.17 which is greater than 1.96 also suggests platykurtic distribution.
Heba Ramadan , I suspect that the reported kurtosis is "excess kurtosis", where the 3 has already been subtracted. (But you'd have to check with the software documentation). ... That still yields a |z| > 2. ... So, now what ?
- Badges
- Science topic
- Similar topics
- Biological Science
- Biota
- Eukaryota
- Fungi