摘要:

Likelihood-based methods are developed for the analysis of field-performance studies with particular attention centered on the estimation of regression coefficients in parametric models.Failure-record dataare those in which the time to failure and the regressor variables are observed only for those items that fail in some prespecified follow-up or warranty period (0,T]. It is noted that for satisfactory inference about baseline failure rates or regression effects it is usually necessary to supplement the failure-record data either by incorporating specific prior information aboutxor by taking a supplementary sample of items that surviveT○. General methods are outlined and specific formulas for various likelihood-based methods are obtained when the failure-time model is exponential or Weibull. In these models the methods are compared with respect to asymptotic efftciency of estimation. Several extensions to more complicated sampling plans are considered.

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488429

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488430

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488431

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488432

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

摘要:

Different copies of a repairable machine, or units, often exhibit different failure rates. If this heterogeneity is ignored, a statistical model of the time until failure may estimate a spurious decreasing hazard rate, resulting in incorrect inferences concerning the risk following repair. This article develops a renewal-proces model that accommodates both heterogeneity across units and decreasing hazard rates. Failure times for each unit are assumed Weibull, and the Weibull scale parameter is assumed to vary across units according to a gamma distribution. The model is illustrated using both Proschan's air-conditioner data and data on a U.S. Navy radar.

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488433

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

摘要:

The expanding use of experimental design techniques for statistical tolerancing is primarily due to their simplicity. They can be understood easily and implemented by engineers and scientists having only a limited knowledge of statistics and experimental design. The method is generally attributed to Taguchi (1978). In this article. we describe Taguchi's method and why it works. both intuitively and mathematically. Our results show that Taguchi's method, although giving good results for many applications, is not optimal. We propose alternative tolerancing procedures that are uniformly better than Taguchi's method with little sacrifice in simplicity. We illustrate the use of these methods. first with some simple examples and then with a partial-differential-equation model.

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488434

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

摘要:

A three-class procedure for acceptance sampling by variables is introduced as an alternative to both the three-class attributes plans introduced by Bray, Lyon. and Burr (1973) and the two-class variables plans introduced by Jennett and Welch (1939). The procedure, which requires that the quality characteristic he normally distributed. has the advantage of requiring a smaller sample size than a three-class attributes plan with approximately the same operating characteristic (OC) surface. The advantage of the three-class variables plan over the two-class variables plan is an ability to discriminate between lots with high and low percentages of marginally conforming items. Two equivalent methods of stating the decision rules for the plans are suggested. It is shown that the OC surfxe for the three-class variables procedure may be constructed bith a special case of the bivariate noncentraltdistribution. Motivation and methods for choosing a plan are discussed and illustrated.

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488436

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

摘要:

When some variables are quantitative and some qualitative in a response-surface context, standard designs may not be suitable. The reasons for this are illuminated. Some alternative designs are discussed. Everyn-point design providesnlinearly independent estimation contrasts. Some of these,psay, are needed to estimate thepparameters of the postulated model. The remainingn—plinearly independent estimation contrasts are available to estimate pure error (if used) and to test for lack of fit, either overall or in particular ways. The key to choosing a good design is to use the available degrees of freedom well, given certain assumptions about the model to be fitted. When there is also uncertainty about the model assumptions, dogmatic design advice is not possible. Sound guidelines are available, however, and these are presented and illustrated.

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488437

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

摘要:

Existing algorithms for the construction of exactD-optimum designs customarily search a list of candidate points. This article describes a computationally simple method of improving such designs. Tables are given of designs for second-order models with as many as five factors found by search over candidate points. These are compared with designs already in the literature and used as the starting point for the adjustment algorithm.

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488438

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor

摘要:

We extend the diagnostic method of Cook and Wang (1983) for assessing case influence to regression models in which both response and linear predictor are transformed. We also suggest a possible improvement of Cook and Wang's method.

ISSN:0040-1706    

DOI:10.1080/00401706.1988.10488439

出版商:Taylor & Francis Group    

年代:1988

数据来源: Taylor