Abstract
This contribution addresses the problem of parameter estimation from an incomplete Poisson sample, i.e. from the part of the random sample where zero values are missing. The aim is to estimate the size of the original sample $N$ and the parameter $\lambda$ of the Poisson distribution.
We focus on the estimators from two papers. The newer one claims that, in the relevant part, it only reviews the procedure from the older paper.
However, it uses different likelihood functions and arrives at slightly different estimators. Derivations are missing in both of the papers and hence we present detailed derivations here and we clarify the connections between the papers.
This contribution is based on the bachelor thesis of the first author.