R help - Aug 2010 - Unequal variance ANOVA using gls function in nlme

Hi

I am trying to run an ANOVA on data with unequal variance.  I am new to nlme,
but to my understanding I need to use the gls function.  I have single response
variable (distance which is continuous) and the explanatory variable is
individual ID (class variable: individuals differ in the variance in their
distance values hence the need to using nlme).

So I would create a model Distance~Individual ID and then create a variance
structure to add to the model.  However the variance structure is not a linear
relationship, it is simply a series of variances linked to individual ID.  Can
anyone guide me as to how to fit this into the model?  As the model is very
simple, I think that the variance is the only component I need to estimate for
the model, but perhaps I am mistaken?

Thanks

Sam


Dr Samantha Patrick
EU INTERREG Post Doc
Davy 618
Marine Biology & Ecology Research Centre
University of Plymouth
Plymouth
PL4 8AA

T: 01752 584699
M: 07740472719


	[[alternative HTML version deleted]]