assessing occuTTD model fit

Skip to first unread message

Christina Aiello

Sep 8, 2022, 2:23:27 PMSep 8
to unmarked

I'm wondering if anyone had recommendations for assessing the fit of time to detection models using either unmarked or ubms. There are fewer examples and info for these models since they haven't been part of the packages for very long. I'm not 100% sure how to best approach the censored observations and what type of fit statistic would be best to use. I'd like to evaluate a fairly simple model - single season, with 1 covariate on psi and  3 on lambda. 

The published studies I've read using ttd models run their models and gof stats in JAGS or similar language, and the gof statistics seem to vary. I'm a novice in using and interpreting that language, so I'd like try and find a way to approach this with unmarked or ubms if possible. Apologies if this is too broad a question but thank you to anyone willing to share some thoughts or references to look to.

Reply all
Reply to author
0 new messages