Abstract
The problem of estimating the probability of a rare event when the sample shows no incidence of the event is considered. Several methodologies based on various statistical techniques are described and their relative performances are investigated. A decision theoretic approach for estimation of response probability when the sample contains zero responses is examined in depth. The properties of each method are discussed and an example from teratology is used to provide illustration and to demonstrate the results.
DOI
10.22237/jmasm/1036110000
Included in
Applied Statistics Commons, Social and Behavioral Sciences Commons, Statistical Theory Commons
