Proper way to estimate a 3-way transition by subtraction

questions concerning analysis/theory using program MARK

Proper way to estimate a 3-way transition by subtraction

Postby Myfanwy » Mon Jan 23, 2017 7:58 pm

I've only ever specified linear arrays before, so I apologize for what is likely a very basic question about transition probability specification in Program MARK. I work with fish, and I've set up a multi-state model to estimate survival of my fish in one of 3 routes as they migrate down a river (routes A, B, and C).

The point of transition is from Route A, where after a certain point a fish must do one of three things:

Stay in Route A (PsiAA)
Transition from A - B. (PsiAB)
Transition from A - C. (PsiAC)

A fish cannot go between route B and C, and both routes B and C eventually converge back to A. I've set up the PIM so that these transition probability parameters are set to either 0 or 1 at the appropriate encounter intervals.

My question is, how do I estimate PsiAA by subtracting both PsiAB and PsiAC from 1? When I open "Change PIM Definition", my three stratas pop up, but obviously I can only estimate a single Psi parameter in Route A by subtraction, and not two.

Would setting up a "dummy route" "D" where the transition of PsiAD = 0 is fixed, but I could then estimate PsiDB or PsiDC through subtraction (which would actually equal PsiAB and PsiAC, respectively) be a solution?

Thank you, please let me know if I have omitted any required information to help answer the question.
Myfanwy
 
Posts: 5
Joined: Mon Jan 23, 2017 3:34 pm

Re: Proper way to estimate a 3-way transition by subtraction

Postby jlaake » Mon Jan 23, 2017 9:48 pm

So unless I'm totally misunderstanding you, what you want is what MARK does by default. It estimates the probability of transitioning from A to B and from A to C and computes A to A by 1-PsiAB-PsiAC. There is nothing you need to do to specify that unless you chose a different stratum to be computed by subtraction.
jlaake
 
Posts: 1417
Joined: Fri May 12, 2006 12:50 pm
Location: Escondido, CA

Re: Proper way to estimate a 3-way transition by subtraction

Postby Myfanwy » Tue Jan 24, 2017 6:01 pm

Well that is very good news, thank you! So then I can calculate the overall probability of taking route A by subtracting the other estimated parameters from 1?

Ex: Parameter estimates from my model:

Psi A to B 0.4505739
Psi A to C 0.4508842

Psi A to A = (1 - Psi A to B - Psi A to C) = 0.0985

(This probability seems very small for the number of fish that actually did stay in Route A, but I'm still working out the kinks in the model parameterization).
Myfanwy
 
Posts: 5
Joined: Mon Jan 23, 2017 3:34 pm


Return to analysis help

Who is online

Users browsing this forum: Google [Bot] and 11 guests