A Bayesian analysis of multiple-recapture sampling for a closed population

摘要: SUMMARY This paper considers from a Bayesian viewpoint inferences about the size of closed animal population data obtained by multiple-recapture sampling scheme. The method developed enables prior information and catch probabilities to be utilized produce considerable improvements in certain cases on ordinary maximum likelihood methods. Several ways expressing such are explored practical example uses these is given. main result an approximation posterior distribution sample that exhibits contributions made ideas. scheme involves taking samples animals, at each stage counting number marked animals sample, marking previously unmarked returning population. literature reviewed Cormack (1968) papers most relevant our problem those Chapman (1952) Darroch (1958). In papers, estimates for parameters variances Our approach differs previous ones introducing size, propensities captured, called 'catch probabilities' some authors. incorporation feature not shared many approaches. (1958) found their introduction into his model does change size. will show strong knowledge can greatly affect inference it this respect greatest difference approaches lies. Although recognized models open incorporating death immigration as Jolly (1965) Seber more practically realistic, we feel method, which deals only with population, worth investigating step towards providing treatment problem.