The great book says on 11-4
in order to use individual covariate data, you
must include the encounter history for each individual in the data file – you can’t summarize your data
by calculating the frequency of each encounter history as you may have done earlier (see Chapter 2 for
the basic concepts if you’re unsure). Each line of the .INP file contains an individual encounter history.
The encounter history is followed immediately by a single digit ‘1’, to indicate that the frequency of this
individual history is 1 (or, that each line of data in the .INP file corresponds to 1 individual).
However, I assume this is just a matter of practical function for most cases and not a mathematical necessity. Because, I assume with data cloning, we would just change that 1 to 100 (for example) and run the same model.
Ditto same question, but for time-varying individual covariates.
I had some unexpected results using RDHuggins data where all the detection estimates were super wonky, so I'm just chasing this down. I used time-varying binary size data as a function of p. Thanks!