Total time on test processes and applications to failure data

Abstract : This paper describes new methods for analyzing nonnegative observations. In particular, data plots are constructed that help in model identification and provide information about the failure rate function. The plots allow the analysis of incomplete data. Properties of the plots are discussed and it is explained how to use them in data analysis and in testing for exponentiality. Examples using real data are given that illustrate the useage of the plots. The asymptotic expression of the process induced by the plots is obtained and it is explained how to use it in generating confidence bands and in testing hypotheses. (Author)