Thanks for all contributions to my question on normalising FA 0 inflated data.
 
After reading your suggestions it seems that normality distribution can be 
skipped, even if it is being portrayed as important in fluctuating asymmetry 
data treatment.
 
A clever point came from James Roper who suggested just exculding 0 data, 
which practically means symmetry of a treat; while our interest is to see the 
variability of asymmetry between traits/habitats. Nevertheless, even after 0 
removal, and other further adjustments such as removing false directional 
asymmetry and trying different transformations the distribution was not normal 
for all traits/sites we considered .  We therefore opted for 
nonparametric models as they can do just as fine. 
Dragos Zaharescu


Animal Anatomy Lab
Faculty of Biological Sciences
Vigo University, apd. 137
36310, Vigo, SPAIN
[email protected]
[email protected]
http://webs.uvigo.es/zdragos/
 
~ You should be the change you want to see in the world ~ Ghandi




________________________________
From: Adam T. Ford <[email protected]>
To: Dragos Zaharescu <[email protected]>
Sent: Tue, March 30, 2010 9:29:37 PM
Subject: Re: [ECOLOG-L] normalization of zero inflated distribution

Hello Dragos,

I would use a Zero Inflated Poisson (ZIP) distribution with a generalized 
linear model in R.

There is some great, accessible information in Zuur et al. mixed model book.

Good luck,

Adam


On Tue, Mar 30, 2010 at 12:03 PM, Dragos Zaharescu <[email protected]> 
wrote:

Is there any way of normalizing 0 inflated da
>
>Dear colleagues,
>
>Is there any way of normalizing 0 inflated data ?
>
>We work on insect fluctuating asymmetry (small, random deviations from 0 of 
>bilateral traits size) which for some traits show too many 0 (though this is 
>not bad).
>Any comments are appreciated.
>
>Thanks 
>Dragos G. Zaharescu
>
>
>Animal Anatomy Laboratory
>Faculty of Biological Sciences
>Vigo University, apd. 137
>36310, Vigo (Pontevedra), SPAIN
>[email protected]
>[email protected]
>http://webs.uvigo.es/zdragos/
>Dragos
>
>
>
>Dragos Zaharescu
>Animal Anatomy Laboratory
>Faculty of Biological Sciences
>Vigo University, apd. 137
>36310, Vigo (Pontevedra), SPAIN
>[email protected]
>[email protected]
>http://webs.uvigo.es/zdragos/
>
>~ You should be the change you want to see in the world ~ Ghandi
>
>
>
>




Reply via email to