[R-sig-eco] R-sig-ecology Digest, Vol 44, Issue 10

[Brian Inouye]

Hi Lara,

As Gavin said, aov, lm, and glm are all fitting exactly the same model 
(when using the defaults for each), you are just getting different 
information from the default summary commands.

If you want "type III" SS for your ANOVA table, use the command 
Anova(model, type = 3) inside the 'carr' package.  (It will also provide 
"type II" SS if you want them).  The default for aov() or anova() is the 
"type I" SS table.  Contrary to what often appears in Google searches, 
using drop1() will *not* give you the "type III" SS -- those are more 
like "type II", in that R follows the marginality principle and keeps 
interaction terms last.  As Chris said, you're on your own to decide 
which of those SS flavors you want.

However, since you mention you're using clutch size as the response, I'm 
skeptical that any of the options you are using is appropriate.  Clutch 
size is probable better modeled with a discrete distribution such as 
Poisson or negative binomial, not a normal distribution.  You should 
probably consider a GLM with Poisson distribution, check for 
overdispersion, and then ask for more help if you've got overdispersion 
problems.

-Brian Inouye

On 11/11/2011 6:00 AM, r-sig-ecology-request at r-project.org wrote:
> >  On 11/11/2011, at 10:25, "Lara R. Appleby 04"
> >  <Lara.R.Appleby.04 at alum.dartmouth.org>   wrote:
> >
>> >>     I'm trying to basically do a two way ANOVA on the dependent variable (clutchsize)
>> >>  with the two independent variables (treatment and species). It seems that there
>> >>  are three ways I can say this in R:
>> >>
>> >>     1. glm(clutchsize~treatment*species)
>> >>     2. aov(clutchsize~treatment*species)
>> >>     3. anova(lm(clutchsize~treatment*species)
>> >>
>> >>     Methods 2 and 3 yield equivalent results, but Method 1 is completely different!
>> >>
>> >>     Any idea why?
>> >>
>> >>     Lara Appleby
>> >>