Summary
The problem of allocating a single observation to one of the two available populations is considered. Suppose that a certain characteristic has densityf in one population, and has densityg in the other. On the basis of the value observed, one must specify which population has densityf. It is assumed that when a wrong population is chosen, a certain known loss is incurred. The problem is to allocate the observation so as to minimize the expected loss. General conditions onf andg are derived to decide which population should be selected for taking the observation.
Similar content being viewed by others
References
Abramson, L. R. (1966). Asymptotic sequential design of experiments with two random variables,J. Roy. Statist. Soc., B,28, 73–87.
Blackwell, D. (1951). Comparison of experiments,Proc. Second Berkeley Symp. Math. Statist. Prob., 93–102.
Bradt, R. N. and Karlin, S. (1956). On the design and comparison of certain dichotomous experiments.Ann. Math. Statist.,27, 390–409.
Chernoff, H. (1972).Sequential Analysis and Optimal Design, SIAM, Philadelphia.
DeGroot, M. H. (1970).Optimal Statistical Decisions, McGraw-Hill, New York.
Feldman, D. (1972). Some properties of Bayesian orderings of experiments,Ann. Math. Statist.,43, 1428–1440.
Lehmann, E. L. (1959).Testing Statistical Hypotheses, John Wiley, New York.
Author information
Authors and Affiliations
Additional information
Research was supported in part by the National Science Foundation under grant No. SOC79-06386.
Work done while on sabbatical leave at Carnegie-Mellon University.
About this article
Cite this article
Gulati, C.M. Selection of certain dichotomous experiments. Ann Inst Stat Math 32, 421–431 (1980). https://doi.org/10.1007/BF02480346
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/BF02480346