Prediction intervals for Weibull order statistics

Using a conditional method, explicit formulae for computing quantiles pertinent to prediction intervals for future Weibull order statistics are developed for two cases: when only previous independent failure data are available, and when both previous independent failure data and early-failure data in current experiment are available. The second case includes the case when only current early-failure data are available. Comparisons of interval widths are made for different estimators of parameters and different ways of forming prediction intervals