Title :
A Useful Property of ``Graphical´´ Estimators of Location and Scale Parameters
Author_Institution :
National Council of Research, Naples, CNR-Istituto Motori-Sez. Statistica; Piazza Barsanti e Matteucci 1; 80125 Napoli ITALY.
Abstract :
This paper presents a method for constructing exact s-confidence limits for reliability and/or quantiles based on graphical estimators of location and scale parameters. The method arises from the ancillary property of the reliability function estimated by using plotting positions and the least squares method. From the distribution of this estimator alone, one can obtain graphs or tables to calculate directly the s-confidence limits for both reliability and quantiles. These limits are exact even for very small sample-sizes in that they are calculated without using asymptotic approximations. An example of a graph which enables one to calculate these limits for a 2-parameter Weibull distribution, is given. This example uses plotting positions recently proposed by the author.
Keywords :
Councils; Least squares approximation; Least squares methods; Maximum likelihood estimation; Parametric statistics; Reliability engineering; Reliability theory; State estimation; Statistical distributions; Weibull distribution; Ancillary statistic; Least squares method; Plotting positions; s-Confidence limits; s-Tolerance limits;
Journal_Title :
Reliability, IEEE Transactions on
DOI :
10.1109/TR.1981.5221125