On Analyzing Mixed Samples
Sleeba John · Journal of the American Statistical Association · 1970
This article describes the application of the method of moments and the maximum likelihood method to identify the population of origin of each observation when it is thought that the data are a mixture of random observations from two different populations. Specifically, four cases in which both the populations belong to the binomial, the negative binomial, the Poisson or the hypergeometric family are discussed. Both methods also yield estimates of the parameters of the populations involved. The asymptotic sampling distributions of the moment estimators are given, and the methods are illustrated by an example.