Some limit results in estimation of proportion based on group testing (Q2338102)

In group testing, each individual subject has a binary response (e.g., positive or negative). It is assumed that the binary responses of the individuals are independent and identically distributed. Instead of testing each of the individuals, groups formed by $k$ individuals are tested in order to estimate the proportion $p$ of individuals with positive response. In the paper under review, two sampling schemes for implementing group testing are considered: one is binomial sampling in which the number of groups $n$ is predetermined, and another one is negative binomial sampling where the total number $n$ of groups with a trait is predetermined. The author proposes both the frequentist and Bayesian estimators of $p$ (the performance of these estimators depends on $n$ and $k$). For the Bayesian estimators it also depends on the hyper-parameter in the prior Beta distribution of the proportion $p$ and describes their asymptotic behavior as $n$ or $k$, or the hyper-parameter tends to infinity.