Page 1 of 1

parametric bootstrapping for goodness-of-fit

PostPosted: Sun Oct 06, 2019 11:22 pm
by donna
Hi all,

I'm using RMark to run Barker and CJS models on my dataset, and would like to test my models for goodness-of-fit using parametric bootstrapping. I've found a good example of code for the dipper dataset that involves analysing the data using groups here: https://sites.google.com/site/cmrsoftwa ... essing-fit

My dataset does not involve groups, and I have tried modifying the dipper code to analyse without groups (so that I can then apply to my dataset) but am clearly getting something wrong somewhere. I'm wondering if anyone could point me in the direction of some existing code to perform parametric bootstrap gof tests without grouping the data.

Would really appreciate help with this - I am a very late stage PhD student with only a few weeks until I submit my thesis. This is literally the last thing I have left to finish and I've been going around in circles with it forever!

Thanks so much!

Re: parametric bootstrapping for goodness-of-fit

PostPosted: Tue Oct 08, 2019 11:42 am
by jlaake
You may want to contact the person who wrote the material you referenced. But as the material shows you can also export to MARK from RMark and use the MARK bootstrap facilities or the median c-hat test or use RELEASE results if appropriate.

--jeff

Re: parametric bootstrapping for goodness-of-fit

PostPosted: Wed Oct 09, 2019 1:25 am
by donna
Thanks for the response, Jeff.

I actually did try exporting and running the simulations through MARK, but the status bar just hung with no progress for ages and I gave up. Having not used MARK before (only RMark), I wasn't sure how long I should expect the simulations to take. I might try leaving it to run overnight and see if anything happens :)