christophe tourenq wrote:
>
> I am a biologist (definitely not a statistician) in France with very
> limited access to statistical resources and I wonder if someone can give
> me some advice. I have been conducting weekly bird counts in wetland
> areas for a period of 1 year. The intention is to compare numbers of
> birds (of each species) among different wetland types. I was intending
> on using a generalized linear model framework where counts are treated
> as a Poisson random variable. The problem is this:
>
> Of course I have now been fully advised that these counts are not
> independent... and someone suggested to me to consider this as a
> repeated measure. However, it seems to me that independence is really
> also a function of time. Clearly counts taken over a short time
> interval are related... but counts taken during different seasons over
> the year probably are not. Someone else suggested to me that there
> might be a way to use a bootstrapping approach to resample the data and
> eliminate some of the independence problem. Anyway, I would appreciate
> any ideas that anyone might have.
The counts are non-independent if you have repeated counts of the same wetland
area (not type) over time. There are models for repeated measures Poisson GLMs
(e.g., multilevel approaches). A normal repeated measures model (possibly with
a transformation) might fit the data reasonably well and is definitely worth
considering. If different areas are sampled (randomly or in a stratified) over
time then a standard Poisson model should be OK.
Thom
