The attached R object contains 3 data objects created by sampling from the same population with 2 groups (males and females) over 10 occasions: (1) a joint live-dead data set, (2) live recaptures only, (3) dead recoveries only.
Save the object to your working directory
From within an R job load the data with the command
load("joint.data")
For each data type process the data and run a series of models using RMark. E.g., you can process the data and run the Phi(dot)p(dot) CJS model
live.processed=process.data(data=joint.data$live,model="CJS")
live.ddl=make.design.data(live.processed)
live.mod1<-mark(live.processed,live.ddl)
Create models that involve group, time, and combination of group and time for each parameter type in the data structure
Collect the models, and compute model average estimates of survival and (if present) fidelity
Compare the estimates of similar parameters between different data structure types. Some differ- why? Some have more precision- why?
Repeat the joint data model comparison using the MARK GUI. You will see differences in AIC weights and model average estimates. Explain why.