Sequential Life Testing With Underlying Normal And Weibull Sampling Distributions
Free (open access)
D. I. De Souza Jr.
Sequential life testing with underlying normal and Weibull sampling distributions D. I. De Souza Jr. Civil Engineering Department & Industrial Engineering Department, Fluminense Federal University & North Fluminense State University, Brazil Abstract The sequential life testing approach gathers sample information only until there is enough to allow a decision with a desirable degree of confidence. The sample size is a random variable and is determined by the result of the analysis of the observed data. It happens that even with the use of a sequential life testing approach, sometimes the number of items necessary to reach a decision about accepting or rejecting a null hypothesis is quite large, as shown by De Souza . In situations like that, the development of a truncation mechanism is essential to guarantee the major advantage of using sequential life testing; that is, small sample sizes. In this work, we will develop a sequential life testing approach in which the underlying sampling distributions are the normal and the Weibull models. We will use the two underlying models to analyze a life testing situation, comparing the results obtained from both. We will also develop a truncation mechanism for the Weibull and Normal models. We will provide rules to truncate a sequential life testing situation making one of the two possible decisions at the moment of truncation; that is, accept or reject the null hypothesis Ho. An example will develop the proposed truncated sequential life testing approach for the Weibull and Normal models. 1 Introduction The two-parameter Weibull distribution has a shape parameter B, which specifies the shape of the distribution, and a scale parameter 8, which represents the characteristic life of the distribution. Both parameters are positive.