A bootstrap method for estimating bias and variance in statistical fisheries modelling frameworks using highly disparate datasets

Type Article
Date 2014
Language English
Author(s) Elvarsson B. P.1, 2, Taylor L.1, Trenkel VerenaORCID3, Kupca V.4, Stefansson G.1
Affiliation(s) 1 : Univ Iceland, Inst Sci, IS-107 Reykjavik, Iceland.
2 : Marine Res Inst, IS-121 Reykjavik, Iceland.
3 : IFREMER, Nantes, France.
4 : Umea Univ, High Performance Comp Ctr North HPC2N, Umea, Sweden.
Source African Journal Of Marine Science (1814-232X) (Natl Inquiry Services Centre Pty Ltd), 2014 , Vol. 36 , N. 1 , P. 99-110
DOI 10.2989/1814232X.2014.897253
WOS© Times Cited 6
Note Online supplementary material: The technical details of the model are available in Supplementary Appendix A at http://dx.doi.org/10.2989/1814232X.2014.897253
Keyword(s) bootstrapping, correlated data, fish population dynamics, non-linear models
Abstract Statistical models of marine ecosystems use a variety of data sources to estimate parameters using composite or weighted likelihood functions with associated weighting issues and questions on how to obtain variance estimates. Regardless of the method used to obtain point estimates, a method is required for variance estimation. A bootstrap technique is introduced for the evaluation of uncertainty in such models, taking into account inherent spatial and temporal correlations in the datasets, which are commonly transferred as assumptions from a likelihood estimation procedure into Hessian-based variance estimation procedures. The technique is demonstrated on a real dataset and the effects of the number of bootstrap samples on estimation bias and variance estimates are studied. Although the modelling framework and bootstrap method can be applied to multispecies and multiarea models, for clarity the case study described is of a single-species and single-area model.
Full Text
File Pages Size Access
13 477 KB Access on demand
9 369 KB Access on demand
Author's final draft 29 514 KB Open access
Top of the page