Back To Index Previous Article Next Article Full Text


Statistica Sinica 17(2007), 301-316





EFFECTS OF MEASUREMENT ERROR AND

CONDITIONAL SCORE ESTIMATION IN

CAPTURE-RECAPTURE MODELS


Wen-Han Hwang, Steve Y. H. Huang and C. Y. Wang


National Chung Hsing University, Tamkang University and
Fred Hutchinson Cancer Research Center


Abstract: Although the literature in covariate measurement error is rather rich, the focus is primarily on regression coefficient estimation; far less is known in the context of capture-recapture experiments. In this article, we provide justification for effects of measurement error on estimation in a capture-recapture model. When errors are present, the conventional approach is shown to have bias in parameter estimation and it may underestimate the population size. In fact, no consistent estimation has been proposed before, especially for the functional case. We propose a new conditional score estimation to adjust for measurement error in capture-recapture models. This approach estimates regression parameters and population size consistently without imposing any distributional assumption on the covariates. An example involving the bird species Prinia flaviventris is used to illustrate the effects, and intensive simulation studies are conducted to evaluate the performance of the proposed estimator along with other existing methods. Under most simulation scenarios, this new method is preferable since it has smaller biases and better coverage probabilities.



Key words and phrases: Capture-recapture, conditional score, Horvitz-Thompson, measurement error, population size, regression calibration.

Back To Index Previous Article Next Article Full Text