Interpreting site Lamda values: Royle-Nichols Count model

[pcutter]

I have output from count a model with one site-wise individual covariate and one survey-occasion individual covariate (survey effort).

I am simply trying to figure out at what level of survey effort the Lamdas are reporting for. Is this somewhere in the Presence output?

Thanks!

[darryl]

Entirely depends upon the design matrices you've used. Which parameters are those covariates on?

[pcutter]

Hi Darryl,

Below is the relevant part of the output with the model specified. I am a little confused about the design matrix functionality after one has run a model. Can it be pulled back up for reviewing a particular model or is one left only with the output?

I think it may be that specified this entire model with a matrix for the wrong *type* of model altogether. Is it possible (would be through blatent ineptitude in my case) to have the "Input" screen and the "Design Matrix" be a mismatch so that model results say one model but reflect a weird matrix based on another model?

Thanks for your help.
__________________________
Royle Model w/ species counts (k=200)
Number of groups = 1
Number of sites = 157
Number of sampling occasions = 4
Number of missing observations = 520



Matrix 1: rows=3, cols=5
-,a1,a2,a3,a4,
p 0 distance 1 0
lambda prey 0 0 1
========================
Number of parameters = 4
**** Numerical convergence may not have been reached.
**** Parameter estimates converged to approximately 3.269116
**** significant digits.

Number of function calls = 149
Final function value = 309.515856
-2log(likelihood) = 619.031711
AIC = 627.031711
Naive occupancy estimate = 0.508475

Untransformed (beta) parameters:
Estimated parameter estimate std.err
-------------------------- -------- -------
beta0 = 0.7628 0.0238
beta1 = 0.0761 0.0089
beta2 = -2.8447 0.1583
beta3 = 1.7853 0.0576

beta var-cov matrix:
0.0006 0.0000 -0.0001 -0.0009
0.0000 0.0001 -0.0011 -0.0001
-0.0001 -0.0011 0.0251 -0.0007
-0.0009 -0.0001 -0.0007 0.0033

[darryl]

The design matrix that is included in the output is that one that PRESENCE thinks you're trying to fit to the data. This isn't the one you think you're trying to fit to the data, then you haven't done something right. The best to check a design matrix in PRESENCE is to start setting up a new analysis and just retrieve the model you want to look at.

Lambda is your estimate of 'abundance' at each site. Survey effort doesn't really come into the interpretation of lambda because that's part of the observation process, not what you/PRESENCE thinks is really out there.