Dear Luis and Javier,
I am having a bit trouble understanding the model structure that was posted
in the original question:
glm(formula = Frec ~ Group+ Hour:Category, family = poisson)
I am an ecologist and I recall from all courses on statistical modelling,
that if I find an interaction term significant, I HAVE to retain the
related main effects as well, even if they have no significant effect by
themselves.
So, for my understanding, why is the above model missing the main effects
'Category' and 'Hour' although their interaction is significant?
Maybe somene can post a clarifying message that could explain what I am
missing or misunderstanding here.
As for the post-hoc test within 'Group', I usually use glht() from the
'multcomp' package.
Thank you vey much!
Cheers,
Mango
On 21 September 2013 08:05, Javier Atalah <javier.ata...@cawthron.org.nz>wrote:
> Hi Luis,
> Try the function testInteractions from the package phia.
> Best regards,
> Javier
>
>
> ________________________________________
> From: r-sig-ecology-boun...@r-project.org [
> r-sig-ecology-boun...@r-project.org] on behalf of Luis Fernando García [
> luysgar...@gmail.com]
> Sent: 21 September 2013 10:40
> To: r-sig-ecology@r-project.org
> Subject: [R-sig-eco] Post Hoc tests GLM
>
> Dear Fellows,
>
> I am trying to build a model in order to test the influence of the time of
> the day, on sex and behavioural category. After performing several
> modelations I chosen the best model which is
>
> glm(formula = Frec ~ Group+ Hour:Category, family = poisson)
>
> summarizing I have
>
> Df Deviance Resid. Df Resid. Dev Pr(>Chi)
> NULL 41 6027,36
> Group 2 6,15 39 6021,21 0,0462
> Category:Hour 13 5972,08 26 49,13 <0,0001
>
>
> I found significative effects of sex and behaviour:time interaction.
> Nevertheless I am having lot of troubles in order to make post hoc
> comparison, specially in the case of Group. Does somebody know how to
> perform post hoc comparisons when using GLM, if so I would really
> appreciate it.
>
> Thanks in advance for your help
>
>
> (attached original data)
>
>
> #####################################################################################
>
> Note:
> This message is for the named person's use only. It may...{{dropped:15}}
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology@r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>
[[alternative HTML version deleted]]
_______________________________________________
R-sig-ecology mailing list
R-sig-ecology@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
