ABSTRACT
The development of (maximum) likelihood-based procedures for estimation of
exponential properties is straightforward. This is not the case, however, for the
Weibull and normal distributions, for which we must introduce procedures for
obtaining asymptotically correct Fisher-matrix or likelihood ratio confidence
intervals on various reliability metrics of interest. The results are generalized
for any multiply right-censored data set. We present worked-out examples
illustrating the use of Microsoft Excel or Minitab for likelihood estimation for
each of the distributions discussed.