1 Scope
This document provides guidelines and procedures for evaluating the conformity of an article (entity, object or system) with the requirements of the specification. The item can be (for example) a measuring block, a scale for a food store, or a blood sample. The following procedures can be used in the following cases:
a) The item under test is distinguished by a single scalar (measurable property, see 3.2.1), which is defined in sufficient detail to be represented by the basic only true value.
a) The item under test is distinguished by a single scalar (measurable property, see 3.2.1), which defines a detailed procedure sufficient to be represented by a basic unique truth value;
Note: GB/T 27418-2017 provides a rationale for not using the term "true", but this document will still use the term, otherwise ambiguity or confusion may arise.
b) The allowable value range for the property is defined by one or two tolerance limits;
c) the property is measured, and the measurement results (see 3.2.5) are expressed in the manner of GB/T 27418-2017, so that the property's
value of the knowledge can be passed;
1) probability density function (PDF, see 3.1.3); 2) distribution function (see 3.1.2);
3) numerical approximation of these functions;
4) the best estimation of juice with inclusion intervals and corresponding inclusion probabilities declared in a reasonable manner.
The procedure proposed in this document applies to the interval that produces the allowable change in the measured value of the attribute of interest, called the acceptance interval. Acceptance limits can be reasonably selected to effectively balance the risk of accepting non-conforming items (consumer risk) or the risk of rejecting conforming items (producer risk).
This document addresses two kinds of conformity assessment issues. The first is to set acceptance limits to ensure that the desired probability of conformity of a single article under test is achieved; the second is to set acceptance limits to ensure that multiple articles measured for toxicity! The second is to set the acceptance limit to ensure that the measurement of multiple items of toxicity (nominal identical) to achieve an acceptable average confidence level. This document provides guidelines for solving this problem.
This document contains examples to illustrate the above guidelines. The concepts presented in this document can be extended to more general conformity assessment problems based on measuring a set of scalars being measured. Some documents, such as references [13, 19], cover industry-specific conformity assessment issues.
This document is intended for use by quality directors, members of standards development organizations and accreditation bodies, as well as testing and calibration laboratories, inspection bodies, accreditation bodies, regulatory authorities, institutions and research institutes.
2 Normative reference documents
The following documents constitute the essential provisions of this document through the normative references in the text. Among them, note the date of the referenced documents, only the date of the corresponding version applies to this document; do not note the date of the referenced documents, the latest version (including all the change orders) applies to this document.
GB/T 3358.1-2009 Glossary of statistical terms and symbols Part 1: General statistical terms and terms used for probability (ISO 3534-1:2006, IDT)
GB/T 3358.2-2009 Glossary of statistical terms and symbols Part 2: Applied statistics (ISO 3534-2:2006, IDT)
GB/T 27000-2006 Vocabulary and general principles of conformity assessment (ISO/IEC 17000:2004,IDT)
GB/T 27418-2017 Measurement uncertainty assessment and representation
GB/T 27419-2018 Measurement uncertainty assessment and representation Supplement 1: Distribution propagation based on Monte Carlo methods
ISO/IEC Guide 99:2007 International vocabulary of metrology - Basic and general concepts and associated terms general concepts and associated terms( VIM)]
3 Terms and definitions
The terms and definitions defined in GB/T 3358.1-2009, GB/T 3358.2-2009, GB/T 27000-2006, GB/T 27419-2018 and ISO/IEC Guide 99:2007 and the following terms and definitions apply to this document.
For definitions from other documents, the note before the source is part of the definition, and the other notes are only applicable to this document.
4 Practice and symbolic representation
5 tolerance limits and tolerance intervals
5.1 Measurement activities in conformity assessment
6 knowledge of the measured
6.1 Probability and information
6.1.1 In the measurement of conformity assessment, a conditional probability density function (PDF) is used to model the knowledge of the attribute of interest (measured), the form of which depends on the information available. This information usually consists of two parts, i.e., information available before the measurement (called a priori information) and more information obtained by the measurement [38].
6.1.2 The probability density function of the attribute of interest (being measured) expresses and conveys its possible values in a given state of knowledge. For conformity assessment specifications, the probability density function of a measure with less information is usually flatter, meaning that the less information is available, the wider the range of possible values. Measurements can obtain new information, steepen the probability density function, and narrow the range of possible values measured.
6.1.3 The role of measurement is to update the (a priori) state of knowledge before measurement to produce the (a posteriori) state of knowledge that contains the measured data after measurement. This transformation rule is called Bayes' theorem, and the underlying mathematical framework is called Bayesian probability theory. The results of this framework are directly adopted in this paper without detailed reasoning or proofs, see references [4,5,16,26,27,39].
7 Probability of compliance with the specified requirements
8 Acceptance interval
8.1 Acceptance limits
8.1.1 The determination of whether an item is accepted as qualified or rejected as unqualified is based on the measured value of the item's attributes nm made and is related to the stated determination rule that specifies the role of measurement uncertainty in the development of acceptance criteria. The interval of the measured value of an attribute that makes the item acceptable is called the acceptance interval (see 3.3.9), and is determined by one or two acceptance limits (see 3.3.8).
8.1.2 As mentioned in the introduction, the choice of the acceptance limits and the corresponding decision rules depends on the severity of the consequences of a miscalculation decision. There are a number of simple and widely used decision rules that are easy to implement. These rules are suitable for situations where the attribute of interest is generalized to the best estimate and the corresponding inclusion interval. Two such rules are described in this chapter.
9 Consumer and producer risk
9.1 General rules
9.1.1 In a conformity assessment using binary determination rules. If the measured value of the article under test is within the defined acceptance interval, the article is judged to be qualified and accepted. If the measured value is outside the acceptance interval, the item is rejected as unacceptable. Figure 12 replicates Figure 1 in the introduction to illustrate the interval of interest, showing the tolerance interval (of the qualified value) and the acceptance interval (of the full measured value).
9.1.2 The use of protective bands limits the probability of making incorrect compliance judgments based on measurement information generalized to the inclusion zone surface. This chapter provides a more precise assessment of this probability for the production process. The evaluation of the probability depends on two factors, i.e. the production process and the measurement system.
9.1.3 If the measurement system is absolutely accurate, all conformity determinations are correct and all risks are zero. An increase in measurement uncertainty means an increase in the probability of misjudgment, which is greatest when the measured value is close to the tolerance limit.
9.1.4 Risk also depends on the characteristics of the production process. If the production process rarely produces items with properties of interest near the tolerance limits, the probability of making a misclassification is low. Conversely, if the process produces items with properties that may be close to the tolerance limits, then the uncertainty associated with the measurement comes into play. The rest of this chapter shows how to assess the contribution of these two factors.
Appendix A (Informative) Compilation of Basic Symbols
Appendix B (informative) Prior Knowledge of the Measured
Appendix C (Informative) Normal Distribution
Bibliography
1 Scope
2 Normative reference documents
3 Terms and definitions
4 Practice and symbolic representation
5 tolerance limits and tolerance intervals
6 knowledge of the measured
7 Probability of compliance with the specified requirements
8 Acceptance interval
9 Consumer and producer risk
Appendix A (Informative) Compilation of Basic Symbols
Appendix B (informative) Prior Knowledge of the Measured
Appendix C (Informative) Normal Distribution
Bibliography
1 Scope
This document provides guidelines and procedures for evaluating the conformity of an article (entity, object or system) with the requirements of the specification. The item can be (for example) a measuring block, a scale for a food store, or a blood sample. The following procedures can be used in the following cases:
a) The item under test is distinguished by a single scalar (measurable property, see 3.2.1), which is defined in sufficient detail to be represented by the basic only true value.
a) The item under test is distinguished by a single scalar (measurable property, see 3.2.1), which defines a detailed procedure sufficient to be represented by a basic unique truth value;
Note: GB/T 27418-2017 provides a rationale for not using the term "true", but this document will still use the term, otherwise ambiguity or confusion may arise.
b) The allowable value range for the property is defined by one or two tolerance limits;
c) the property is measured, and the measurement results (see 3.2.5) are expressed in the manner of GB/T 27418-2017, so that the property's
value of the knowledge can be passed;
1) probability density function (PDF, see 3.1.3); 2) distribution function (see 3.1.2);
3) numerical approximation of these functions;
4) the best estimation of juice with inclusion intervals and corresponding inclusion probabilities declared in a reasonable manner.
The procedure proposed in this document applies to the interval that produces the allowable change in the measured value of the attribute of interest, called the acceptance interval. Acceptance limits can be reasonably selected to effectively balance the risk of accepting non-conforming items (consumer risk) or the risk of rejecting conforming items (producer risk).
This document addresses two kinds of conformity assessment issues. The first is to set acceptance limits to ensure that the desired probability of conformity of a single article under test is achieved; the second is to set acceptance limits to ensure that multiple articles measured for toxicity! The second is to set the acceptance limit to ensure that the measurement of multiple items of toxicity (nominal identical) to achieve an acceptable average confidence level. This document provides guidelines for solving this problem.
This document contains examples to illustrate the above guidelines. The concepts presented in this document can be extended to more general conformity assessment problems based on measuring a set of scalars being measured. Some documents, such as references [13, 19], cover industry-specific conformity assessment issues.
This document is intended for use by quality directors, members of standards development organizations and accreditation bodies, as well as testing and calibration laboratories, inspection bodies, accreditation bodies, regulatory authorities, institutions and research institutes.
2 Normative reference documents
The following documents constitute the essential provisions of this document through the normative references in the text. Among them, note the date of the referenced documents, only the date of the corresponding version applies to this document; do not note the date of the referenced documents, the latest version (including all the change orders) applies to this document.
GB/T 3358.1-2009 Glossary of statistical terms and symbols Part 1: General statistical terms and terms used for probability (ISO 3534-1:2006, IDT)
GB/T 3358.2-2009 Glossary of statistical terms and symbols Part 2: Applied statistics (ISO 3534-2:2006, IDT)
GB/T 27000-2006 Vocabulary and general principles of conformity assessment (ISO/IEC 17000:2004,IDT)
GB/T 27418-2017 Measurement uncertainty assessment and representation
GB/T 27419-2018 Measurement uncertainty assessment and representation Supplement 1: Distribution propagation based on Monte Carlo methods
ISO/IEC Guide 99:2007 International vocabulary of metrology - Basic and general concepts and associated terms general concepts and associated terms( VIM)]
3 Terms and definitions
The terms and definitions defined in GB/T 3358.1-2009, GB/T 3358.2-2009, GB/T 27000-2006, GB/T 27419-2018 and ISO/IEC Guide 99:2007 and the following terms and definitions apply to this document.
For definitions from other documents, the note before the source is part of the definition, and the other notes are only applicable to this document.
4 Practice and symbolic representation
5 tolerance limits and tolerance intervals
5.1 Measurement activities in conformity assessment
6 knowledge of the measured
6.1 Probability and information
6.1.1 In the measurement of conformity assessment, a conditional probability density function (PDF) is used to model the knowledge of the attribute of interest (measured), the form of which depends on the information available. This information usually consists of two parts, i.e., information available before the measurement (called a priori information) and more information obtained by the measurement [38].
6.1.2 The probability density function of the attribute of interest (being measured) expresses and conveys its possible values in a given state of knowledge. For conformity assessment specifications, the probability density function of a measure with less information is usually flatter, meaning that the less information is available, the wider the range of possible values. Measurements can obtain new information, steepen the probability density function, and narrow the range of possible values measured.
6.1.3 The role of measurement is to update the (a priori) state of knowledge before measurement to produce the (a posteriori) state of knowledge that contains the measured data after measurement. This transformation rule is called Bayes' theorem, and the underlying mathematical framework is called Bayesian probability theory. The results of this framework are directly adopted in this paper without detailed reasoning or proofs, see references [4,5,16,26,27,39].
7 Probability of compliance with the specified requirements
8 Acceptance interval
8.1 Acceptance limits
8.1.1 The determination of whether an item is accepted as qualified or rejected as unqualified is based on the measured value of the item's attributes nm made and is related to the stated determination rule that specifies the role of measurement uncertainty in the development of acceptance criteria. The interval of the measured value of an attribute that makes the item acceptable is called the acceptance interval (see 3.3.9), and is determined by one or two acceptance limits (see 3.3.8).
8.1.2 As mentioned in the introduction, the choice of the acceptance limits and the corresponding decision rules depends on the severity of the consequences of a miscalculation decision. There are a number of simple and widely used decision rules that are easy to implement. These rules are suitable for situations where the attribute of interest is generalized to the best estimate and the corresponding inclusion interval. Two such rules are described in this chapter.
9 Consumer and producer risk
9.1 General rules
9.1.1 In a conformity assessment using binary determination rules. If the measured value of the article under test is within the defined acceptance interval, the article is judged to be qualified and accepted. If the measured value is outside the acceptance interval, the item is rejected as unacceptable. Figure 12 replicates Figure 1 in the introduction to illustrate the interval of interest, showing the tolerance interval (of the qualified value) and the acceptance interval (of the full measured value).
9.1.2 The use of protective bands limits the probability of making incorrect compliance judgments based on measurement information generalized to the inclusion zone surface. This chapter provides a more precise assessment of this probability for the production process. The evaluation of the probability depends on two factors, i.e. the production process and the measurement system.
9.1.3 If the measurement system is absolutely accurate, all conformity determinations are correct and all risks are zero. An increase in measurement uncertainty means an increase in the probability of misjudgment, which is greatest when the measured value is close to the tolerance limit.
9.1.4 Risk also depends on the characteristics of the production process. If the production process rarely produces items with properties of interest near the tolerance limits, the probability of making a misclassification is low. Conversely, if the process produces items with properties that may be close to the tolerance limits, then the uncertainty associated with the measurement comes into play. The rest of this chapter shows how to assess the contribution of these two factors.
Appendix A (Informative) Compilation of Basic Symbols
Appendix B (informative) Prior Knowledge of the Measured
Appendix C (Informative) Normal Distribution
Bibliography
Contents of GB/T 27430-2022
1 Scope
2 Normative reference documents
3 Terms and definitions
4 Practice and symbolic representation
5 tolerance limits and tolerance intervals
6 knowledge of the measured
7 Probability of compliance with the specified requirements
8 Acceptance interval
9 Consumer and producer risk
Appendix A (Informative) Compilation of Basic Symbols
Appendix B (informative) Prior Knowledge of the Measured
Appendix C (Informative) Normal Distribution
Bibliography