Using hidden Markov models to deal with availability bias on line transect surveys.

We develop estimators for line transect surveys of animals that are stochastically unavailable for detection while within detection range. The detection process is formulated as a hidden Markov model with a binary state-dependent observation model that depends on both perpendicular and forward distances. This provides a parametric method of dealing with availability bias when estimates of availability process parameters are available even if series of availability events themselves are not. We apply the estimators to an aerial and a shipboard survey of whales, and investigate their properties by simulation. They are shown to be more general and more flexible than existing estimators based on parametric models of the availability process. We also find that methods using availability correction factors can be very biased when surveys are not close to being instantaneous, as can estimators that assume temporal independence in availability when there is temporal dependence.

An Introduction to Model Selection.

This paper is an introduction to model selection intended for nonspecialists who have knowledge of the statistical concepts covered in a typical first (occasionally second) statistics course. The intention is to explain the ideas that generate frequentist methodology for model selection, for example the Akaike information criterion, bootstrap criteria, and cross-validation criteria. Bayesian methods, including the Bayesian information criterion, are also mentioned in the context of the framework outlined in the paper. The ideas are illustrated using an example in which observations are available for the entire population of interest. This enables us to examine and to measure effects that are usually invisible, because in practical applications only a sample from the population is observed. The problem of selection bias, a hazard of which one needs to be aware in the context of model selection, is also discussed. Copyright 2000 Academic Press.

Model selection and the estimation of odds ratios in the presence of extraneous factors.

This paper deals with model selection and the estimation of odds ratios from cross-classified frequencies in the presence of extraneous factors. The odds ratio is estimated in different ways dependent on whether the extraneous factor is modelled as an effect modifier, a confounder, or neither. Routinely this choice is based on statistical tests of null hypotheses. By contrast, we propose selection of the model which is estimated to maximize the accuracy of the estimator of the odds ratio, on average. We demonstrate how a non-parametric bootstrap method can be used to carry out the selection, and illustrate the methodology using an example on use of oral contraceptives and myocardial infarction.

On the statistical analysis of ROC curves.

We introduce a new accuracy index for receiver operating characteristic (ROC) curves, namely the partial area under the binormal ROC graph over any specified region of interest. We propose a simple but general procedure, based on a conventional analysis of variance, for comparing accuracy indices derived from two or more different modalities. The proposed method is related to and compared with existing methodology, and is illustrated by results from an experiment on optimization of density and contrast yielded by multiform photographic images used for scintigraphy.