A general class of recapture models based on the conditional capture probabilities. Issue 1 (10th September 2015)

• Record Type: Journal Article
• Authors: Farcomeni, A.
• Is Part Of: Biometrics. Volume 72:Issue 1(2016)
• Page Start: 116
• Page End: 124
• Publication Date: 2015-09-10
• Subjects: Aitchinson‐Silvey algorithm -- Capture history -- Equality constraints -- Heterogeneity -- Population size; Biometry -- Periodicals; 570.15195
• DOI: 10.1111/biom.12375
• Languages: English
• ISSNs: 0006-341X

Abstract

Summary: We propose an M hotb model for population size estimation in capture‐recapture studies. The tb part is based on equality constraints for the conditional capture probabilities, leading to an extremely rich model class. Observed and unobserved heterogeneity are dealt with by means of a logistic parameterization. In order to explore the model class, we introduce a penalized version of the likelihood. The conditional likelihood and penalized conditional likelihood are maximized by means of efficient EM algorithms. Simulations and two real data examples illustrate the approach.