Comparing exponential location parameters with several controls under heteroscedasticity

A note on the multiple comparisons of exponential location parameters with several controls under heteroscedasticity

“…In this section, we present the results of simulation studies in which we compare our two-sided one-stage SCIs in terms of CP and AV with the ones that were proposed by [12], [5], and [10]. We use abbreviations M11, SFGCI, and PBSCI for these three methods.…”

“…Maurya et al [12] showed that each of the four groups of data is exponentially distributed. Maurya et al [12] and Malekzadeh et al [10] previously analyzed this data set. Our two-sided SCIs for comparing the test drugs with the control drugs are shown in Table 4.…”

“…Malekzadeh et al [10] proposed parametric bootstrap simultaneous confidence intervals (PBSCIs) for θ i − µ j , i = 1, . .…”

“…They showed theoretically that their SCIs have correct CP asymptotically; however, they did not investigate the CP of PBSCIs for small samples theoretically. Malekzadeh et al [10] also compared the CP of PBSCIs with SFGCIs via simulation studies and concluded that PBSCIs outperform the SFGCIs for sample size as small as 10.…”

“…In Section 5, we discuss constructing SCIs for comparing the treatment groups with the best control. In Section 6, via a simulation study, we compare the proposed SCIs with the ones that were proposed by [12], [5], and [10] in terms of CP and average volume (AV). Simulation results show that the proposed SCIs have higher CP and smaller AV in most cases.…”