The operating characteristic of double sampling plans by variables when the standard deviation is unknown

Summary We deal with double sampling plans by variables for a one-sided specification limit when the quality characteristic is normally distributed with unknown standard deviation. An algorithm is presented that allows to calculate the OC of the sampling plans proposed by Bowker and Goode (1952). We give several examples. Furthermore, it is shown that the algorithm carries over to calculating the OC of the double-stage t-test. The authors wish to thank Yvonne K?llner and Timor Saffary for technical support.     

Nonparametric covariate adjustment for receiver operating characteristic curves

The accuracy of a diagnostic test is typically characterized using the receiver operating characteristic (ROC) curve. Summarizing indexes such as the area under the ROC curve (AUC) are used to compare different tests as well as to measure the difference between two populations. Often additional information is available on some of the covariates which are known to influence the accuracy of such measures. The authors propose nonparametric methods for covariate adjustment of the AUC. Models with normal errors and possibly non‐normal errors are discussed and analyzed separately. Nonparametric regression is used for estimating mean and variance functions in both scenarios. In the model that relaxes the assumption of normality, the authors propose a covariate‐adjusted Mann–Whitney estimator for AUC estimation which effectively uses available data to construct working samples at any covariate value of interest and is computationally efficient for implementation. This provides a generalization of the Mann–Whitney approach for comparing two populations by taking covariate effects into account. The authors derive asymptotic properties for the AUC estimators in both settings, including asymptotic normality, optimal strong uniform convergence rates and mean squared error (MSE) consistency. The MSE of the AUC estimators was also assessed in smaller samples by simulation. Data from an agricultural study were used to illustrate the methods of analysis. The Canadian Journal of Statistics 38:27–46; 2010 © 2009 Statistical Society of Canada     

Characterization of MIL-STD-105D double sampling plans using normalized OC curves

A simple method of analyzing the MIL-STD-105D double sampling plans by ‘normalizing the OC curve’ is presented. Among a set of competing double sampling plans, procedures are given to select a particular plan when costs are unknown. Necessary tables are constructed and examples given.     

Local linear smoothing of receiver operating characteristic (ROC) curves

For measuring the accuracy of a continuous diagnostic test, the receiver operating characteristic (ROC) curve is often used. The empirical ROC curve is the most commonly used non-parametric estimator for the ROC curve. Recently, Lloyd (J. Amer. Statist. Assoc. 93(1998) 1356) proposed a kernel smoothing estimator for the ROC curve and showed his estimator has better mean square error than the empirical ROC curve estimator. However, Lloyd's estimator involves two bandwidths and has a boundary problem. In addition, his choice of bandwidths is ad hoc. In this paper we propose another kernel smoothing estimator which involves only one bandwidth and does not have the boundary problem. Furthermore, our choice of the bandwidth is asymptotically optimal.     

A test for crossing receiver operating characteristic (roc) curves

The receiver operating characteristic (ROC) curve gives a graphical representation of sensitivity and specificity of a prediction model when varying the decision treshold on a diagnostic criterion. A classical test for comparing the overall accuracies for two models -1 and 2- is based on the difference between ROC curves areas - related to its standard error. This test is designed for the situation where ROC curve 1 caps ROC curve 2. Often both curves cross :in this paper, a new test, based on the integrated difference between the curves, is proposed to deal with this situation. In a simulation experiment, the new test was less powerful than the old test for detecting an overall superiority, but much more powerfull against the crossing alternative.     

Limits of experiments associated with sampling plans

Starting from Milbrodt (1985), the asymptotic behaviour of experiments associated with Poisson sampling, Rejective sampling and its Sampford-Durbin modification is investigated. As superpopulation models so-called L^r-generated regression parameter families (1⩽r⩽2) are considered, allowing also the presence of nuisance parameters. Under some assumptions on the first order probabilities of inclusion it can be shown that the sampling experiments converge weakly if the underlying shift parameter families do so. In case of convergence the limit of the sampling experiments is characterized in terms of its Hellinger transforms and its Lévy-Khintchine representation, leading to criteria for the limit to be a pure Gaussian or a pure Poisson experiment respectively. These results are then applied to the situation of sampling in the presence of random non-response, and to establish local asymptotic normality (LAN) under more restrictive conditions. Applications also include asymptotic optimality properties of tests based on Horvitz-Thompson-type statistics, and LAM bounds and criteria for adaptivity, when testing or estimating a continuous linear functional in LAN situations. They especially cover the case of sampling from an unknown symmetric distribution, which has been subject to detailed investigations in the i.i.d. case.     

Three-stage chain sampling plans

This article outlines the structure of a generalized family of three-stage chain sampling plans, extending the concept of two-stage chain sampling plans of Dodge and Stephens (1966) which is an extension of the original work of Dodge (1955). Expressions are derived for the OC curves for several sets of general sets of three-stage chain sampling plans, with cumulative acceptance numbers of (c₁,c₂,c₃) = (0,1,2), (0,1,3), (0,2,3), (1,2,3),(0,1,4),(0,2,4), (0,3,4), (1,2,4), (1,3,4) and (2,3,4). In the original work of Dodge (1955) only acceptance numbers of 0,1 were used and in the extension work of Dodge and Stephens (1966) acceptance numbers of (c₁,c₂) = (0,1) ,(0,2), (1,2), (0,3), (1,3), (0,4) and (1,4) were used with selected sets of values of k₁,and k₂ (the number of lots considered for cumulation in the first and second stage respectively). In this paper the 0C curves are derived more generally for any k₁, k₂and k₃(the number of lots considered for cumulation in the first, second and third stages respectively) combination and 0C curves of a number of plans are given and comparisons are made with some single sampling, two-stage chain sampling and multiple sampling plans.     

Rank statistics expressible as integrals under P–P-plots and receiver operating characteristic curves

Given a probability measure on the unit square, the measure of the region under an empirical P – P -plot defines a two-sample rank statistic. Instances include trimmed and censored versions of the Mann–Whitney–Wilcoxon statistic and a class of statistics with applications in the analysis of receiver operating characteristic (ROC) curves. A large sample distribution for such a statistic is obtained, which is valid under sampling from general populations. Explicit results are presented for comparing arbitrary quantile segments of two populations. The results are not restricted to continuous data and incorporate adjustments for tied values in the discrete case. A multivariate version of the large sample distribution extends the class of tractable statistics in ROC analysis and facilitates the use of methods based on partial areas when the data are discrete.     

Exact two-sided Lieberman-Resnikoff sampling plans

We deal sith sampling by variables with two-way-protection in the case of aN(μσ²) distributed characteristic with unknown σ². For the sampling plan by Lieberman and Resnikoff (1955), which is based on the MVU estimator of the percent defective, we prove a formula for the OC. If the sampling parametersp ₁ (AQL),p ₂ (LQ) and α, β (type I, II errors) are given, we are able to compute the true type I and II errors of the usual (one-sided) approximation plans. Furthermore it is possible to compute exact two-sided Lieberman-Resnikoff sampling plans.     

Acceptance sampling plans for linear profiles with one-sided specifications

In this study, we propose three new sampling plans based on yield index C_puA for linear profiles with one-sided specifications, including the resubmitted sampling plan, the repetitive group sampling plan, and the multiple dependent state repetitive sampling plan. The operating characteristic functions of our proposed sampling plans are developed. The plan parameters of our proposed sampling plans are determined through nonlinear optimization. The plan parameters are reported for various combinations of acceptable quality level and limiting quality level. The three sampling plans are compared with the existing single sampling plan in terms of the average sample number. A real example is used to illustrate the applications.     

Variable repetitive group sampling plans with process loss consideration

Traditionally, most acceptance sampling plans considering the fraction defective do not distinguish among the products that fall within the specification limits. However, products that fall within the specification limits may not be good if their mean is far away from the target. So, developing an acceptance sampling plan with process loss consideration is essential. In this paper, a variable repetitive group sampling plan is proposed to deal with process loss. The design parameters of the proposed plan are tabulated for various combinations of acceptance quality levels. The proposed methodology can be used to determine whether the products meet the desired levels of protection for both producers and consumers.     

A note on the operating characteristic curve of the known sigma single sampling variables plan

The operating characteristic curves of certain known sigma variables sampling plans may not be satisfactory in that they have a tendency to reject even lots of acceptable quality. This note presents the theory and a method to identify such known sigma variables plans possessing unsatisfactory operating characteristic curves.     

Modified tightened two-level continuous sampling plans

In this paper, a modification is proposed on the tightened two-level continuous sampling plan. The tightened two-level plan is one of the three tightened multi-level continuous sampling plans of Derman et al. (1957) with two sampling levels. A modified tightened two-level continuous sampling plan is considered, for which the rules concerning partial inspection depend, in part, on the length of time it takes to decide that the process quality is good enough that 100% inspection may be suspended (e.g. the time required to find i consecutive items free of defects). Using a Markov chain model, expressions for the performance measures of the modified MLP-T-2 plan are derived. The modified MLP-T-2 plan is shown to be identical to the MLP-T-2 plan. Tables are also presented for the selection of the modified MLP-T-2 plan when the AQL or LQL and AOQL are specified.     

Inverse-gaussian acceptance sampling plans by variables

Variable sampling plans to control fraction defective are obtained using the Inverse-Gaussian (IG) distribution. OC curves are obtained and impact of sample size and specification limits on these curves are discussed. Simulation studies are used to investigate sensitivity of the sampling plans under the more commonly used normal distribution.     

ASN-minimax double sampling plans for variables

X be a continuous quality characteristic, with one-sided lower specification limit, having either normal distribution with known σ or exponential distribution. We report on an algorithm allowing the calculation of the so-called ASN-Minimax plan. This plan has minimal maximal average sample size among all double sampling plans for variables that obey the classical two-points-condition on the operating characteristic. We give examples and tables of the ASN-Minimax plans in the normal as well as in the exponential case. Received: March 20, 2000; revised version: January 22, 2001     

Simulated and unbiased reliability sampling plans

Statistical estimation and hypothesis testing are two ways in which the data of a sample can be made to yield information concerning the parameters of the population from which the sample is drawn. The most important applications of these two ways is the acceptance sampling methods used in industrial quality control, systems reliability and failure detection. The use of these sampling methods is connected with the risk of unnecessary rejection of satisfactory lots and the risk of acceptance of lots with defective units, these decisions can occur when selecting biased samples containing most defective units or no defective units despite the fact that the lot has the contrary. The aim of this research is to make this kind of decision an improbable or impossible event. The parameters and techniques determining the sampling methods must be correctly chosen. The formulation of an optimised statistical model of these problems is the basic condition of obtaining objective results. By the essence of statistical simulation the process of functioning of the complex system was used to represent a mathematically formulated model which was isomorphic in all essential aspects of the research objectives. This model was repeatedly tested to determine the required statistical characteristics, based on the complex stochastic process.     

Generalized tightened two-level continuous sampling plans

In 1955, Lieberman and Solomon introduced multi-level (MLP) continuous sampling plans. Derman et al . then extended the multi-level plans as tightened multi-level plans (MLP-T). In this paper, a generalization of MLP-T with two sampling levels is presented. Using a Markov chain model, expressions for the performance measures of the general MLP-T plans are derived. Tables are also presented for the selection of general MLP-T plans with two sampling levels when the acceptable quality level, limiting quality level, indiff erence quality level and average outgoing quality level are specified.     

Selection of CSP-V continuous sampling plans

MIL-STD-1235C establishes standard procedures for the selection and implementation of single- and multi-level continuous sampling plans, such as CSP-1, CSP-F, CSP-2, CSP-T and CSP-V. CSP-V is a single-level continuous sampling procedure which provides for alternating sequences of 100%inspection (either at full or reduced clearance number) and sampling inspection. It requires a return to 100% inspection whenever a non-confirming unit is found during sampling inspection, but provides for a reduced clearance number upon demonstration of superior product quality. The CSP-V procedure serves as an alternative to the CSP-T procedure where a reduction in sampling frequency has no economic merit. In this paper, expressions for the average outgoing quality, the average fraction inspected and the operating characteristic function are derived using a Markov chain model. Four tables are given to enable the selection of CSP-V plans when the acceptable quality level or the limiting quality level and the

average outgoing quality limit are specified.     

Quick specification of acceptance sampling plans for dependent attribute sampling

Traditionally, fraction-nonconforming is treated as a Bernouilli variable with the attendant assumption of independence in constructing the acceptance sampling plans. This study relaxes the independence assumption in attribute sampling and provides a convenient nomogram to incorporate the effects of dependence into the sampling plans. The nomogram provides convenient answers for a number of questions about sampling with attributes. The study simplifies the procedural details, virtually eliminates the computation and facilitates the rapid scanning and evaluation of alternative solutions.     

Improved attribute acceptance sampling plans based on maxima nomination sampling

This paper demonstrates the use of maxima nomination sampling (MNS) technique in design and evaluation of single AQL, LTPD, and EQL acceptance sampling plans for attributes. We exploit the effect of sample size and acceptance number on the performance of our proposed MNS plans using operating characteristic (OC) curve. Among other results, we show that MNS acceptance sampling plans with smaller sample size and bigger acceptance number perform better than commonly used acceptance sampling plans for attributes based on simple random sampling (SRS) technique. Indeed, MNS acceptance sampling plans result in OC curves which, compared to their SRS counterparts, are much closer to the ideal OC curve. A computer program is designed which can be used to specify the optimum MNS acceptance sampling plan and to show, visually, how the shape of the OC curve changes when parameters of the acceptance sampling plan vary. Theoretical results and numerical evaluations are given.