Communications in Statistics - Theory and Methods, Volume 43, No. 23, Year 2014
Notification
URL copied to clipboard!
In dose-response studies, Wadleys problem occurs when the number of organisms that survive exposure to varying doses of a treatment is observed but the number initially present is unknown. The unknown number of organisms initially treated has traditionally been modelled by a Poisson distribution, resulting in a Poisson distribution for the number of survivors with parameter proportional to the probability of survival. Data in this setting are often overdispersed. This study revisits the beta-Poisson distribution and considers its effectiveness in modelling overdispersed data from a Wadleys problem setting.