Model averaging of beta estimates

[Shawn Craik]

I have 3 nest survival models with some degree of support. Two covariates are included in all 3 models and the other covariates are only included in one or two of these models. I was wondering if it is possible for MARK to model average the beta estimates of the covariates from these 3 models to account for any uncertainty in their effects on nest survival?
Thanks,
Shawn

[abreton]

Rerun each covariate model that will be included in the model averaging exercise but before you do, see the options to use 'Mean Individual' or 'User-specified Covariate Values' on the 'Setup Numerical Estimation Run' window in MARK. As long as you use the same values for equivalent covariates in each model then the model averaged probabilities will make sense. MARK does not offer a 'beta' model averaging option (you'd have to do this on your own in, e.g., Excel...).