saduria at hotmail.com wrote:> I'd like to compare length-weight regressions among years. Any
information would be appreciated.
>
> a. gray
> fisheries consultant
Your message is rather cryptic for a general statistical audience,
whereas I'm sure in a fisheries group everybody would understand what
you want.
Use a glm with family=Gaussian(link=log) for all data sets together (in
original units) with year as a factor, then run the model again ignoring
the year factor, and compare the different fits using anova(name of your
glm objects) for a likelihood ratio test, or inspect the AICs for a
non-frequentist model selection method.
R.