ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979341

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

Easy to compute one-sided small sample conservative prediction intervals for binomial, Poisson, hypergeometric, and negative-binomial distributions are described. These are obtained using a procedure that connects prediction interval problems to those which arise in the ranking and selection of statistical populations. As a result, known tables can be used to find prediction factors. Further, factors for prediction intervals based on a method suggested in the literature are given for the binomial and Poisson cases.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979342

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

An optimal design of CUSUM control charts is reviewed. Plots of chart parameters are given which enable the chart parameters of an optimal CUSUM chart to be determined easily. Both the CUSUM charts with and without head starts are considered.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979343

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

Focusing on the process, not the product, is the key to implementing statistical process control in low-volume manufacturing environments. One approach is to initially assume that all products are being produced by a common process. Data can be used to verify or refute this assumption. If it is refuted, the appropriate subprocesses that must be monitored separately can be identified. A case study of a small batch machining process is used to describe this approach to implementing low-volume SPC.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979344

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

Approximately normalized control charts, called Q Charts, have been given by the author in other work for charting variables and for a binomial attributes chart. We consider here the corresponding transformations for Poisson distributions, and give the Q charts technology for controlling a Poisson parameterλ. Charts are given for the two cases when the Poisson parameterλis known and when it is unknown. The chart for the case whenλis unknown is useful both for short runs and for charting processes at start up before enough data are available to assume thatλis essentially known. The chart for the case when the parameter is assumed known gives probabilities of exceeding the control limits that are better approximated by the nominal normal distribution probabilities than the classicalcorucharts. We give some tabular results to compare the goodness of the normal approximation to the probabilities for these charts with thecorucharts, and also with a chart made using a square root normalizing transformation.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979345

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

A“one-shot”device has the property that a successful test results in its destruction. An obvious example is the testing of an explosive device selected from a stockpile of military weapons. The type of data obtained differs from that of the standard life-testing situation. Such data is dichotomous (go or no go) rather than data obtained by measuring a continuous variable such as failure time or stress level. The emphasis in this paper is on test plans for periodic testing of highly reliable one-shot devices. Based on model assumptions, the sequence of minimal sample sizes, achieving specified reliability criteria, are derived.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979346

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

We consider the economic selection of both the process mean and the upper control limit in a two-level process-control scheme for a capacitated (bottleneck) production process. A closed-form expression for the optimal upper control limit is developed, and a one-way table and an approximating equation are provided for the optimal mean. The paper quantifies the savings in expected profit relative to using a single level of process control, and identifies the economic conditions under which the second level of process control is most beneficial.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979347

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

The repeated uniform filling of a package and the subsequent accurate estimation of its contents are complicated by the variability of the fill and packaging materials. Consequently, although each container is weighed, the process will still yield underfilled and overfilled packages. Therefore, screening limits are employed as a countermeasure, beyond which the product is reclaimed. The lower and upper limits are designed to decrease the risk of noncompliance and minimize product“giveaway,”respectively. In this paper, a control policy model is developed for this environment which determines the economically optimal settings of the mean and the screening limits, subject to an acceptable level of risk.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979348

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

The well-known model of Grubbs, which is based on repeatability conditions, is generalized to reproducibility conditions in an interlaboratory test study. ANOVA estimators are developed for the repeatability and reproducibility variances of each procedure as well as for the variability within and between laboratories. The paper gives a solution to a problem which often occurs in practice, but which cannot be tackled by the standard model of ISO 5725.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979349

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor

摘要:

A new variables control chart for location based on the Hodges-Lehmann estimator associated with the Wilcoxon signed rank statistic is presented. The proposed chart is nonparametric and hence maintains the nominal type I error specified. Calculations of control limits are simplified for certain conditions. The proposed approach compares favorably with the Shewhart control chart when the data are from a Gaussian distribution. Its performance is better than the traditional approach in the case of moderate sample sizes from long-tailed symmetric distributions. These properties make it well suited for early production runs of limited size when the distribution of the process statistic is unknown. The method may also be used to establish robust control limits.

ISSN:0022-4065    

DOI:10.1080/00224065.1991.11979350

出版商:Taylor&Francis    

年代:1991

数据来源: Taylor