« first day (2367 days earlier)      last day (2936 days later) » 

Ell
Ell
08:55
Hi - I'm using GLMMs (and mixed effects cox regressions in another case) to test for significant effects on a response variable. I have fixed effects of sex and some environmental variables, and I am interested in testing for significant sex by environment interactions (are responses to the environment sex-specific?) so the full models will be y~sex*factor1*factor2*factor3+(random1)+(random2)...
does anyone have advice on material to read about testing for these, and how to go about model selection when I have an interaction of particular interest? I've only worked on random effects and explained variation previosuly
09:35
@Ell I'd ask that on the main site. This is mostly for gossip and grumbles; it's not a more informal version of CV itself. Direct questions get seen by many more people and are much more durable.
 
3 hours later…
Ell
Ell
12:27
@NickCox I thought of asking here as it is quite basic/broad, perhaps not suitable for the site - more looking for direction to resources where I can work out the answer and understand why it is the answer :)
12:52
I am just giving standard advice here in chat. Technical questions belong in the main forum. There are parts of SE where trying out questions informally is encouraged, but this room in CV really is just focused on chat. Basic questions don't belong here any more than any others and broad questions can't be answered at all well.
2
You have two or three questions all at once, so the telegraphic (dare I say Twitter-like) style couldn't cope. Even people who could answer wouldn't want to spend time writing something that will just sink into oblivion. Do ask on CV!
3
Ell
Ell
thanks

« first day (2367 days earlier)      last day (2936 days later) »