[R-sig-eco] Offsets in Poisson or Neg. Bin regression

[Highland Statistics Ltd]

Matias,
The only problem with the offset is that you implicitly assume:

double the number of embryos per individual (FECUND)  ==> double the 
expected value of damaged embryos per individual

This simply follows from the equation that Philip wrote down. For some 
scenarios this makes sense, but not for other scenarios.

Alain

> Matias,
>
> The situation when you most want to use an offset is when FECUND differs across individuals.  If FECUND is a constant for all observations, you could ignore it if you chose to.  If you did that, its constant effect gets rolled into the intercept.
>
> When it's not constant, the logic is:
>
> log(mu_i / F_i) = your model   <==> Log(mu_i) = your model + log(F_i)
> where mu_i is the mean count and F_i is the fecundity, both for individual i.
>
> When F_i is a constant = F, the logic is:
> log(mu_i / F) = your model   <==> Log(mu_i) = B0 + rest of your model + log(F)  <==> B0 + log(F)  + rest of your model
> where B0 + log(F) is the new "intercept"
>
> Best wishes,
> Philip Dixon
>
>
>
> ------------------------------
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>
>
> End of R-sig-ecology Digest, Vol 63, Issue 12
> *********************************************
>


-- 

Dr. Alain F. Zuur
First author of:

1. Analysing Ecological Data (2007).
Zuur, AF, Ieno, EN and Smith, GM. Springer. 680 p.
URL: www.springer.com/0-387-45967-7


2. Mixed effects models and extensions in ecology with R. (2009).
Zuur, AF, Ieno, EN, Walker, N, Saveliev, AA, and Smith, GM. Springer.
http://www.springer.com/life+sci/ecology/book/978-0-387-87457-9


3. A Beginner's Guide to R (2009).
Zuur, AF, Ieno, EN, Meesters, EHWG. Springer
http://www.springer.com/statistics/computational/book/978-0-387-93836-3


4. Zero Inflated Models and Generalized Linear Mixed Models with R. (2012) Zuur, Saveliev, Ieno.
http://www.highstat.com/book4.htm

Other books: http://www.highstat.com/books.htm


Statistical consultancy, courses, data analysis and software
Highland Statistics Ltd.
6 Laverock road
UK - AB41 6FN Newburgh
Tel: 0044 1358 788177
Email: highstat at highstat.com
URL: www.highstat.com
URL: www.brodgar.com