Goodness-Of-Fit Test for MSORD model

questions concerning analysis/theory using program MARK

Goodness-Of-Fit Test for MSORD model

Postby emilylester » Tue Jun 02, 2015 9:03 pm

Hello,

I have run a MSORD model in MARK and am wondering which goodness-of-fit test I can run on this model.

15.8 of the book tells me to run a Pearson chi-squared test in ORDSURVIV, which sounds good but I cannot find any information on how to run this program (and it does not look very user friendly).

Is there a way that I can run this test in MARK?

Thanks!
Emily
emilylester
 
Posts: 7
Joined: Thu May 14, 2015 12:32 am

Re: Goodness-Of-Fit Test for MSORD model

Postby cooch » Tue Jun 02, 2015 9:52 pm

emilylester wrote:Hello,

I have run a MSORD model in MARK and am wondering which goodness-of-fit test I can run on this model.

15.8 of the book tells me to run a Pearson chi-squared test in ORDSURVIV, which sounds good but I cannot find any information on how to run this program (and it does not look very user friendly).


Stepping for Bill and Jim (who I assume will jump in and correct any errors -- I'm working from memory here, which is getting less adaptive as I get older).

Actually, ORDSURVIV is very user-friendly - and far simpler than MARK, since it comes with ~6 pre-built models pre-selected. All you need to do is feed it the correctly formatted input file (which is similar, but not identical to what MARK uses).

1\ Just grab the ordsurviv.zip from the Patuxent Software Archive, extract it the 3 files, and then double-click the executable called ordsrv.exe. [Note: If you're using a Mac without Windows installed, you're SOL.]

2\ This will bring up a little GUI which allows you to select the .inp file (use the turt_hist.dat example -- have a look to see how the encounter data are formatted), specify number of primary sampling periods, and secondaries per sampling period, and then

3\ its not much harder than running the pre-defined models (click the convert/gen models button).

4\ Thats about it. If you're only interested in GOF, I believe you simply select the most general of the pre-defined models (model 3, if I recall correctly), and run that. Notepad will pop open, with lots of information about how the model is parameterized, the estimates, and, at the bottom, all the component Pearson chi-sq tests.

5\ You can consider overall chi-sq/df as a rough estimate of c-hat (similar to what you do with RELEASE, which is discussed in some detail in Chapter 5). The Pearson tests in ORDSURVIV are, I believe, essentially equivalent to those in 'regular' SURVIV, which has extensive documentation in various places. So, you can always look there.

Is there a way that I can run this test in MARK?


No.
cooch
 
Posts: 1628
Joined: Thu May 15, 2003 4:11 pm
Location: Cornell University

Re: Goodness-Of-Fit Test for MSORD model

Postby emilylester » Wed Jun 03, 2015 5:08 am

THANK YOU SO MUCH!
emilylester
 
Posts: 7
Joined: Thu May 14, 2015 12:32 am

Re: Goodness-Of-Fit Test for MSORD model

Postby violetblue » Mon Jan 07, 2019 1:04 am

Does anyone have any thoughts on using R2ucare functions for the GOF of an MSORD model?

I'm thinking of possibly pooling the secondary into the primary occasions and continuing as for the geese Arnason-Schwarz example in the R2ucare documentation.

Or should I still continue with ORDSURVIV as described above?
violetblue
 
Posts: 23
Joined: Tue Nov 06, 2018 11:16 pm


Return to analysis help

Who is online

Users browsing this forum: Google [Bot] and 12 guests