Failing to take into account varying catchability of individuals in capture-recapture models is a cause of correlation among sources. Models have been developed for that random components, mainly in ecological experiments where researchers control the capture-mark-recapture technology. In the epidemiological studies the probelm is more complex, as individuals are heterogeneous and the researcher does not control the capture mechanism. After a brief review of the methodology underlying capture-recapture experiments, we focus on the problem underlying epidemiological studies. We propose a class of models that take into account both heterogeneity of individuals and correlations among sources, when no covariates are available. We apply it to the estimation of diabetes in a town of Northern Italy.
Latent class models for varying catchability and correlation among sources in Capture-Recapture estimation of the size of a human population
STANGHELLINI, Elena;
1999
Abstract
Failing to take into account varying catchability of individuals in capture-recapture models is a cause of correlation among sources. Models have been developed for that random components, mainly in ecological experiments where researchers control the capture-mark-recapture technology. In the epidemiological studies the probelm is more complex, as individuals are heterogeneous and the researcher does not control the capture mechanism. After a brief review of the methodology underlying capture-recapture experiments, we focus on the problem underlying epidemiological studies. We propose a class of models that take into account both heterogeneity of individuals and correlations among sources, when no covariates are available. We apply it to the estimation of diabetes in a town of Northern Italy.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
