|
Abstract:
|
Population abundances are rarely, if ever, known. Instead, they are estimated with some amount of uncertainty. The resulting measurement error has consequences on subsequent analyses that model population dynamics and estimate probabilities about abundances at future points in time. This work addresses outstanding questions about the consequences of, and ways to correct for, measurement error in several dynamic models. We provide analytical results on the biases that result from a naive analysis, which ignores measurement error under a class of models that allow for heteroscedasticity in the measurement errors. We also develop and evaluate moment- and likelihood-based methods to correct for measurement error, with and without the use of estimated measurement error variances. The methods are applied to existing population dynamics datasets. This is joint work with John Buonaccorsi.
|
