Constraining temperatures

questions concerning analysis/theory using program MARK

Constraining temperatures

Postby Pau Lucio » Fri Apr 22, 2022 3:44 am

Dear all,

My first post in the forum. I hope you can help me. I have read the Mark book several times but still struggle.

I am working with the annual survival (5 years) of Moustached warbler and Reed Warbler, and my aim is to check the effect of temperatures and precipitation on survival. When I constrain the models with precipitation it works fine but when I do with temperatures and p is time-dependent Phi (Temp) p (t), it shows the warning: Number of parameters estimated from gap method (=5) and threshold method (=6) differ. If I delete the first column with 1 (slope) it works fine. Please could you advise why that is happening and what would be the best approach? Many thanks

https://drive.google.com/file/d/1QZBFSURvvKIfe3JNlM92jkfogGofGQ7Y/view?usp=sharing

PS:I have problems embedding the image so I link it
Pau Lucio
 
Posts: 3
Joined: Thu Apr 21, 2022 3:43 am

Re: Constraining temperatures

Postby cooch » Fri Apr 22, 2022 10:37 am

Pau Lucio wrote:Dear all,

My first post in the forum.


Welcome!

<snip>... it shows the warning: Number of parameters estimated from gap method (=5) and threshold method (=6) differ. If I delete the first column with 1 (slope) it works fine. Please could you advise why that is happening and what would be the best approach? Many thanks

https://drive.google.com/file/d/1QZBFSURvvKIfe3JNlM92jkfogGofGQ7Y/view?usp=sharing


The short answer is that MARK has a number of things running 'in the background' to try to determine the number of parameters that can be estimated, given (i) the model, and (ii) the limits of the data you are fitting the model to. The warning you refer to relates to this issue. The technical details can be found by reading the addendum to Chapter 4 in the book, fairly lengthy subsection starting on p. 74 (near the bottom - 'How does MARK count parameters?'). There is a subsection on issues that might arise when 'threshold does not equal gaps', starting on p. 80.

In general, this sort of issue arises when one or more probability parameters are estimated near either the 0 or 1 boundary.


PS:I have problems embedding the image so I link it


Because embedding images has been 'turned off' (ever since hackers learned to embed malicious code in JPG and PNG files).
cooch
 
Posts: 1628
Joined: Thu May 15, 2003 4:11 pm
Location: Cornell University


Return to analysis help

Who is online

Users browsing this forum: No registered users and 12 guests