摘要:

Inspection errors can have a severe impact on an attributes single-sampling plan for lot acceptance due to misclassification of units as defectives or nondefectives. However, with estimates of the errors, the plan can be adjusted to preserve the desired operating characteristic curve or average outgoing quality limit (AOQL). Allowance is made for a skip-lot plan and a sampling plan for rejected lots that have been screened 100 percent.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978802

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

A method is presented for determining whether a sufficient proportion of a product meets specifications when the available data are subject to reading error. The method is illustrated by an application in which a producer is required to verify that at least 95 percent of the containers of a product provide the net weight indicated on the container label. In this example, errors in reading the filled container weights and random variation in the empty container weights might at first lead one to conclude incorrectly that the proportion of containers that meet the requirement is insufficient.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978803

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

A means of providing a quick summary of the status of a quality assurance program for management and a detailed progress report for those responsible for implementing the program is presented. The purpose is to focus attention on, and establish priorities for, quality planning and implementation both in the initial stages and in the re-evaluation of delayed elements or modified processes. Through the use of flow charts, cause and effect diagrams, cause and effect flow charts, priority evaluation reviews, and quality assurance status evaluations, it is possible to focus attention on important considerations in a timely manner.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978804

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

Some of the important aspects of accelerated-stress life testing such as extrapolation distance and sample size determination are considered. Also addressed are the advantages of applying more than one accelerated stress to induce failure.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978805

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

This article develops an alternative clerical performance measure for work submitted to acceptance sampling. The measure,“expected thruput,”appears more comprehensive than approaches based on either actual output, actual thruput, effective thruput or effective production. Expected thruput considers both an employee's production and quality levels. It also provides a curvilinear increase in credit for batches with linear increases in the sample effective proportion. The model is applicable to situations involving multiple jobs, variable batch sizes and different types of acceptance sampling plans.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978806

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

Important decisions are frequently made from the analysis of product life and field data using various graphical and analytical methods. Simple methods, such as fitting a Weibull distribution to the data, however, may give wrong answers, especially when estimates beyond the data range are desired, under certain conditions. The first part of this paper reviews the major purposes of product life analyses, briefly describes some of the standard methods and indicates some pitfalls associated with the use of simple methods. The hazards are illustrated by the analysis of human mortality and electromechanical device data. The second part of the paper discusses pitfalls resulting from product mixtures, proposes some general models for product life analysis and suggests remedial approaches.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978807

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

A statistical procedure based on weighted least squares regression is derived for making inferences regarding mean values of several random variables when only sums of the random variables and the number of occurrences of each variable contributing to each sum are observed. The estimates of the mean values and their sampling distribution are discussed. The application of the procedure to the monitoring of time standards and an example, which is solved by means of a readily available statistical package, are also presented.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978808

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

The probabilities that the sampling plans presented in the new NBS Handbook 133 detect noncompliance with package weight regulations are presented. The tables of probabilities are designed to be easily applied to specific packages and processes that may be of interest. A brief comparison is made to Handbook 67, the predecessor to Handbook 133.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978809

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

摘要:

Most methods for determining double attribute sampling plans depend on tables generated using the Poisson distribution with specific risk levelsαandβ. This paper presents an algorithm that provides the flexibility to be able to use a hypergeometric, binomial, Poisson or normal distribution and specify any values ofαandβ. The computer program then provides several possible plans for the user to examine and provides plots of the OC, AOQ, ASN and ATI curves for the plan having risks less than or equal to those specified and minimum sample size.

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978810

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1982.11978811

出版商:Taylor&Francis    

年代:1982

数据来源: Taylor