PRESENCE unable to differentiate data

announcements (new versions, changes, bugs, installation problems...) related to program PRESENCE

PRESENCE unable to differentiate data

Postby dmh1454 » Fri Aug 02, 2019 12:53 pm

I am currently working on modeling for occupancy rates of semi-aquatic mammals in treated vs. nontreated sites. For this, I separated the treatment data into two groups, treated and control. For the treated data I put all treated sites as 1 and all control as 0 (vice versa for control groups).

Site Treated Control
1 1 0
2 0 1
3 0 1

The problem is, it appears that PRESENCE is unable to understand the difference between what is treated and what is control. When I ran the models it gave me an AIC of 272.83 for both of them, with the exact same output. I was told to try combining the two together, leaving treated as 1 and control as 0 as such.

Site Management
1 1
2 0
3 0

But it gives me the exact same AIC and does not show a difference in the output. Does anyone know where I made a mistake? Sorry if this has already been answered.
Posts: 11
Joined: Fri Aug 02, 2019 12:35 pm

Re: PRESENCE unable to differentiate data

Postby jhines » Fri Aug 02, 2019 1:55 pm

You can use either of those covariates. The design matrix will depend on which you use. If you use the 1st set of covariates (Treated, Control), the design matrix for occupancy would look like this:

Code: Select all
-,    a1,     a2
psi  Treated Control

Or, using an intercept and "treatment" effect you could use:

Code: Select all
-,    a1,     a2
psi  Treated  1

Or, using the 2nd covarite (Management):

Code: Select all
-,  a1, a2
psi  1  Management

The model with no effect of treatment would have the following design matrix:

Code: Select all
-,  a1
psi  1

If that doesn't solver the problem, I think I would need to see the input and output files.

Jim (
Posts: 596
Joined: Fri May 16, 2003 9:24 am
Location: Laurel, MD, USA

Return to software problems/news

Who is online

Users browsing this forum: No registered users and 1 guest