On estimating the total probability of the unobserved outcomes of an experiment

THE UNOBSERVED OUTCOMES OF AN EXPERIMENT* P.J. Bickel and J.A. Yahav University of California at Berkeley Robbins (1968) considered the problem of estimating the total probability of the unobserved outcomes of an experiment. In this paper we suggest an estimator, based on n trials, and show that under some regularity conditions one can construct asymptotic confidence intervals for the random quantity we look for