significant new feature in MARK | general data bootstrapping

announcements (new versions, changes, bugs, installation problems...) related to program MARK

significant new feature in MARK | general data bootstrapping

Postby cooch » Thu Apr 19, 2018 12:07 pm

Gary has just implemented data bootstrapping for *all* models in MARK. If you don't know what data bootstrapping is, the following is a very brief summary:

The data bootstrap we introduce here provides a method to evaluate whether theoretical variance estimates are valid when the individuals are considered statistically independent, or whether variance inflation procedures are required to account for dependence among (say) siblings.


Non-independence of fates among marked individuals is a fairly big deal, and probably more general than we'd like to think (in a perfect world, all study organisms would be blind, semi-sentient with no memory, and would wander around our study area such that the action of each individual is in no way influenced by actions of other individuals. In other words, our study organisms should be Brownian particles. I don't know where this perfect world is, but a lot of statistical models assume we visit there with some frequency).

More details to come, but a fairly complete introduction is found in Appendix G of the MARK book:

http://www.phidot.org/software/mark/doc ... /app_7.pdf
cooch
 
Posts: 1628
Joined: Thu May 15, 2003 4:11 pm
Location: Cornell University

Return to software problems/news

Who is online

Users browsing this forum: No registered users and 2 guests

cron