Quantifying Precision of Mark-Recapture Estimates Using the Bootstrap and Related Methods

摘要: SUMMARY Bootstrap techniques yield variance estimates under any model for which parameter can be calculated, and are useful in cases where analytic variances not available closed form, or only if more restrictive assumptions made. Here the application of bootstrap to mark-recapture models is discussed. The approach also allows generation robust confidence intervals, extend beyond permissible range itself out-of-range estimates. If an animal population assumed (i.e., no death, birth, migration), two further methods obtaining limits size suggested. first based on a Robbins-Monro search each limit, second applies concept randomisation permutation test. In absence nuisance parameters, both exact apart from Monte Carlo variation limitations imposed by discrete distribution. For second, all possible permutations enumerated, eliminated.