Simultaneous confidence intervals for comparing several exponential location parameters with a control

Abstract

In this paper, two-sided simultaneous confidence intervals, on the lines of Hayter et al. (J Stat Plan Inference 86:81–99, 2000), to compare \(k\) two-parameter exponential populations with a control population in terms of location parameters are proposed, which combine the advantages of one-sided simultaneous confidence intervals and two-sided simultaneous confidence intervals of Bofinger (Aust J Stat 34(1):65–75, 1992). The proposed two-sided simultaneous confidence intervals also maintain the inferential sensitivity of positive directional decision of one-sided simultaneous confidence intervals. Computation of the critical constants of the proposed procedure is discussed and selected critical constants are tabulated. Working and advantages of the proposed procedure are demonstrated with a numerical example.