Comparing survival rates of two species

[Pau Lucio]

Hi!
I am working in paper where I would like to compare the survival value of two different sympatric species of warblers. Please, could you recommend any test to check if the difference between both species is significant?

A second question, I am testing if one of these species have different sex-related survival. The best model did not include sex but the second model with 1.9 Delta QAICc did. Should I do model averaging on only this two models or include the rest of the models which have a "reasonable" weight.

Many thanks for your help

[jhines]

There are a few methods.
1) Program Contrast - https://www.mbr-pwrc.usgs.gov/software/contrast.shtml. Enter the two rates and standard errors or var-cov matrix and get a significance probability.
2) Compare the beta value for the effect of the 2nd species to zero. If beta-1.96*std.err < 0 or beta+1.96*std.err > 0, then significant difference. This assumes design matrix has one species as intercept and other species as "effect" or difference between species.
3) Use likelihood-ratio. Run model with different rate for the 2 species and another model with same rate for both species. Difference in Deviance should be chi-square with 1 degree of freedom. Look up difference in chi-square table.

[Pau Lucio]

Many thanks for the help