Nest Survival and interaction effects

[falconflyer23]

Hello fellow MARK users,

I have a question that I have not found the answer to in the book chapters as well as in my search of this forum under "nest survival" and "interaction terms". My study investigates the nesting success of the Brown Creeper in terms of various logging techniques.

I am interested in how treatment, nest age, and year affect nest survival. More specifically I want to analyze these models:

1. Constant
2. Treatment
3. Nest age
4. Year
5. Treatment + nest age
6. Treatment + nest age + year
7. Treatment + year
8. Neat age + year
9. Treatment + nest age + treatment * nest age
10. Treatment + year + treatment * year

So far I have been able to complete the first 8 by following the chapter 17. However, I have run into a roadblock with incorporating interaction terms into my models. Chapter 17 does not deal with this. I have read chapter 6 but the difference is I am only using 1 group and the design matrix looks very different in the nest survival feature.

To answer this question, I need insight on design matrix please. Thank you.

[bacollier]

FalconFlyer23 (interesting handle btw),

Based on your comments, this may be easier for you to do using multiple groups in the PIMS than to mess with the DM for the models you suggest (you should still try to understand how DM's are structured, but you probably need to read/work some more on this).

If you have nests within each treatment (logging cuts), you can make a group out of each treatment (if you have 4 logging techniques, then you have 4 groups), if you have 2 years, then you could have 2*4=8 groups that represent treatment*year combinations and then you can model nest age within each of those. It will likely be simpler for you given your model set is fairly simple.

<climbing on soapbox>
As an aside, I personally don't ever like to use models that include 2 main effects and then an interaction of those main effects, mainly because if the interaction is the biologically important value, then by definition the interpretation of the main effects is not relevant in and of themselves, and if one posits that main effects are important, then the interaction is not.</getting down off soapbox>