摘要:

Nine different sample/rest-of-lot disposition policies for single and double sampling, along with explicitly developed AOQ and ATI expressions for each—both with and without inspection error—are considered. The nine dispositions include all combinations of scrapping/eliminating defectives/replacing defectives in the sample and scrapping/eliminating defectives/replacing defectives in rejected lots.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980980

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

Nature, origin, and treatment of systematic errors in measurements and calibrations are discussed. It is shown how systematic errors can vary, how random errors under one set of conditions will become systematic errors under another set of conditions and vice versa. Recommendations are made concerning the assignment of values to limits of systematic errors of measurements and standards. Some earlier sources are quoted on the philosophy of measurement errors.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980981

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

Double variables sampling plans and matched single variables plans having two fixed points on the operating characteristic curve are presented. These plans provide minimum average sample number when lot quality is at the producer's quality level.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980982

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

A modified skip-lot sampling plan is presented under which at least one unit is always sampled from a lot. The formulas for the acceptance probability and average sample number using this plan are derived and compared with the formulas for the standard skip-lot plan. The derivation of the formulas using a Markov Chain is presented in the Appendix.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980983

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

To simplify the choice of a single gauge sampling plan used to control a process in which the variation follows a normal distribution with known standard deviation, the acceptance number c is taken as (n- 1)/2, wherenis the sample size. This relationship is always nearly optimal and permits the use of a simple nomogram to find the parameters of a plan that meets the user's requirements.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980984

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

Two charts are presented for instant estimation of the parameters of a doubly or singly truncated normal distribution. They can be readily applied by the general quality control practitioner.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980985

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

Principal components and factor analysis are two techniques that are finding increasing application among quality engineers who are concerned with processes with more than one response variable. In this second part of a three-part series, the discussion of principal components will include such topics as grouped data, correlation matrices, interpretation, simplification and rotation of vectors, along with some applications of these techniques. Part three, factor analysis, will be found in the next issue.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980986

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

A set of USDA acceptance sampling plans based on cumulative sum (cusum) procedures is described. The plans, for inspection by attributes, are indexed by AQL's and apply to continuous production grouped in lots or batches. They are simple to apply and provide better discrimination than single sampling plans. The plans differ from published cusum procedures in that starting values are used and the cusum value is not reset to zero after failure.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980987

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

摘要:

A FORTRAN computer program is presented which provides a number of test statistics and their critical values for the simultaneous pairwise comparison of several treatment means. The different test statistics are applicable in specific situations with respect to the balance in the number of observations available for each treatment mean and the consistency or homogeneity of the standard deviations for each treatment. Descriptions of the various test statistics for each situation are given and examples, illustrating the application of these statistical tests and the computer program to actual process quality data, are presented.

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980988

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1981.11980989

出版商:Taylor&Francis    

年代:1981

数据来源: Taylor