[R-sig-Geo] Normality Test for Large sample

Edzer Pebesma edzer.pebesma at uni-muenster.de
Fri Jun 20 21:02:38 CEST 2008


Dear Pujan,

I believe this is not really the most appropriate list to send your 
question to, as it has little to do with spatial data.

Briefly, with samples the size you have, the smallest deviation from 
_exact_ normality will show up as significant. The real question is what 
that means. Is it a problem that your data come from a distribution that 
is _almost_ normal?
--
Edzer

PUJAN RAJ REGMI wrote:
> Dear list,
> I am trying to perform normality test for my data set. my data set is quite large N>13000 with 4 valriables under study , so When I tried to do normality test for each variable using lillie.test of nortest package the p values is almost zero "2.16e-16". So for all variable the test reject the null hypothesis (p-value the same("2.16e-16").i.e. it doesn't comes from normal distribution which I believe is not true and even this can be seen in histogram plot that fwe variable's plot shows normal distribution. Now I doubted that my approach to test for normality might not be correct becasue I am dealing with large number of samples. So I request you to provide me insight on the other efficient ways in R.
> Thanking you
> Sincerely,
> Pujan
>
> _________________________________________________________________
> Earn cashback on your purchases with Live Search - the search that pays you back!
>
> arncashback
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>




More information about the R-sig-Geo mailing list