SYNOPTIC ABSTRACT
Consider k independent exponential populations (treatments or systems) with different scale and location (possibly unknown) parameters. The goal is to select a subset of k exponential populations which includes the “best” population, i.e. the one with the largest value of the scale parameter. The data are assumed to be: (i) complete, or (ii) incomplete with Type-II censoring. The cases of known and unknown location parameters are treated separately. Tables of critical constants which are necessary to carry out the procedure are developed. Performance of the proposed procedure in comparison to the existing procedure is assessed through simulation study. The proposed procedure is also illustrated with the help of a real life example.