residuals

[nlsanto2]

Hello everybody,
I am trying to evaluate the fit of the POPAN models to my data and I am looking at the residuals plots.
My first question is: are the residuals plot (together with TEST2+TEST3) a reliable way to assess the goodness of fit of POPAN models?
Also on the manual at page 13-19 it says that because individual capture histories were used (in the salmon example on the book), residual plots are not useful to assess the goodness of fit...

Does it mean that if my input file is like this:
0100 1;
0100 1;
0100 1;
.....
i cannot use the residuals plot and if it is like this:
0001 19;
0010 53;
0011 5;
0100 83;
....
I can?

I will be very grateful to anybody who will take the time to answer my question!
Nina

[abreton]

Hello Nina ... I just noticed these additional data/question.

As the text reads (provided in my other response), mark requires this format for res plots,

0001 19;
0010 53;
0011 5;
0100 83;

However, keep in mind the dipper dataset example that I mentioned. Those data (see Dipper.INP file) are structured like,

01100 1 0;
11000 0 1;
etc.

... and the res plot feature seems to produce legitimate results. Recall I guessed that the progam might combine identical encounter histories before plotting when NO individual covariates are included in the dataset. So, for the dippers ... you could get something like,

01100 5 2;
11000 3 1;
etc

... prior to mark integrating the EHs into the res plots.

My guess.

andre