Abstract
Many systems involve progression through a series of distinct stages, such as disease or developmental stages. In ecological studies, often individuals such as small arthropods cannot be marked, so data are collected on cohort development. Multistage models for unmarked cohort data use a distribution for each stage duration and possibly stage-specific mortality rates. We generalize previous models and present computational methods for smoothed maximum likelihood estimation. The general model allows arbitrary distribution assumptions, stage-specific mortality, unobserved stages, and correlations between stage durations using Gaussian copulas. Monte Carlo integration of the stage distributions is used to approximate the probabilities needed for the likelihood. We establish a heuristic smoothing step for the simulated probabilities that yields a smooth approximate likelihood surface. For the case of classic grasshopper cohort data, we demonstrate AIC model selection to determine which among past arbitrary constraints are actually justified by the data. Finally, we demonstrate how estimates of stage distribution parameters depend on the unknown stage correlations.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: Journal of Agricultural, Biological, and Environmental Statistics
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.