www.phidot.org

Hi, I'm trying to work out abundance from my CJS model. I'm having real difficulty understanding what the vcv actually is and how this relates to the N-hat estimates. I've had a look through the book but am still struggling to get my head around what exactly the variance-covariance matrix is actually showing.

Could someone please explain exactly how the vcv relates to n-hat?

Here are my outputs;

Code: Select all

$Nhat
         [,1]     [,2]     [,3]     [,4]     [,5]
[1,] 61.61668 39.93673 51.34723 33.09044 21.67994

$vcv
          [,1]     [,2]      [,3]     [,4]     [,5]
[1,] 14.601864 3.831112  4.925716 3.174350 2.079747
[2,]  3.831112 8.116187  3.192594 2.057449 1.347984
[3,]  4.925716 3.192594 11.347267 2.645292 1.733122
[4,]  3.174350 2.057449  2.645292 6.372135 1.116901
[5,]  2.079747 1.347984  1.733122 1.116901 3.789709

Thanks so much!

I'll let Jeff address your VCV question, but...might be worth reading the following thread before you get too invested in abundance estimates from a simple CJS model:

viewtopic.php?f=1&t=3548&hilit=cjs+abundance

Hey Cooch,
Thanks so much for the reply. I had come across that thread (thanks for the link though) and what I understood is to take a 'pinch of salt' attitude to the estimates
Thanks again!

vcv is the variance covariance matrix. The sqrt of the diagonal estimates are the standard errors of the individual estimates. It gives you the full vcv if you want to combine estimates and it is appropriate. For example if the estimates were separate groups of animals and you wanted the sum and its variance you would sum all of the elements of the vcv and take its square root to get the std error of the total abundance.