Zhang, W. , Price, S. J. and Bonner, S. J. (2021) Maximum likelihood inference for the band-read error model for capture-recapture data with misidentification. Environmental and Ecological Statistics, 28(2), pp. 405-422. (doi: 10.1007/s10651-021-00492-6)
Full text not currently available from Enlighten.
Abstract
Misidentification of animals is a common problem for many capture-recapture experiments. Considerably misleading inference may be obtained when traditional models are used for capture-recapture data with misidentification. In this paper, we investigate the so-called band-read error model for modeling misidentification, assuming that it is possible to identify one marked individual as another on each capture occasion. Currently, fitting this model relies primarily on a Bayesian Markov chain Monte Carlo approach, while maximum likelihood is difficult because there is not a computationally efficient likelihood function available. The Bayesian method is exact but requires considerable computation time. We propose an approximate model for modeling misidentification and then develop a fast maximum-likelihood approach for the approximate model using likelihood constructed by the saddlepoint approximation method. We demonstrate the promising performance of our proposed method by simulation and by comparisons with the Bayesian inference under the original model. We apply the method to analyze capture-recapture data from a population of Northern Dusky Salamanders (Desmognathus fuscus) collected in North Carolina, USA.
Item Type: | Articles |
---|---|
Additional Information: | This work was funded by the Natural Sciences and Engineering Research Council of Canada (Grant Number 43024-2016). |
Status: | Published |
Refereed: | Yes |
Glasgow Author(s) Enlighten ID: | Zhang, Dr Wei |
Authors: | Zhang, W., Price, S. J., and Bonner, S. J. |
College/School: | College of Science and Engineering > School of Mathematics and Statistics > Statistics |
Journal Name: | Environmental and Ecological Statistics |
Publisher: | Springer |
ISSN: | 1352-8505 |
ISSN (Online): | 1573-3009 |
Published Online: | 24 March 2021 |
University Staff: Request a correction | Enlighten Editors: Update this record