www.phidot.org

All,

For known fate analysis.

My data consists of: 8 groups (sub-adult males, adult males, sub-adult females, and adult males for 2 release areas. Seasons are the encounter occasions (acclimation period, dry, and wet). My models consist of a proper assortment of sex, age, release area, and season. All models were created using the Parameter Index Chart. No covariates were used.

Out of the 16 models in my candidate set, the best model was sex x season (AICc weight = >60%, next model was 10%).

I want to correctly get and interpret the beta estimates from a design matrix for model sex x season.
1 = male, 0 = female. B3, B4, and B5 correspond to an indication for acclimation period, dry season, and wet season respectively.

Below is my idea of the design matrix. Am I creating this correctly to get a Beta estimate for the sex effect in each season? More importantly, by running this I am getting 5 estimates, which is correct, but from my knowledge of how this works, I need to be getting the last estimate as a reference (with all blanks for the beta, SE, and CIs).

B1 B2 B3 B4 B5
1 1 1 0 0
1 1 0 1 0
1 1 0 0 1
1 0 1 0 0
1 0 0 1 0
1 0 0 0 1

I hope I haven't confused any of you. Please let me know if this makes sense and you know how I can fix the DM.

Taylor

Hi Taylor.

Sorry for not directly answering your question - I'm not totally clear on the problem so I don't want to give you bad advice. However, if you haven't already read it, ch. 6 in the MARK book, especially section 6.2 on linear models and the design matrix, will probably help you interpret your betas.
Also, Jeff Laake's RMark help file, 'Linear models and capture-recapture analysis in MARK' may be useful, although the examples are done in RMark.

Joe