Standard practice for laboratory bias evaluation using single test result from standard material
Some standard content:
People's Republic of China Entry-Exit Inspection and Quarantine Industry Standard SN/T 1492—2004
Method for evaluating laboratory deviation using single test result from standard material
Standard practice for laboratory deviation evaluation using single test result from standard material2004-11-17 Issued
People's Republic of China
General Administration of Quality Supervision, Inspection and Quarantine
2005-04-01 Implementation
This standard adopts the American ASTL6517 standard by reference. Appendix A of this standard is the normative appendix. This standard is proposed and managed by the National Certification and Accreditation Administration. SN/T1492—2004
This standard is jointly developed by Liaoning Exit-Entry Inspection and Quarantine Bureau of the People's Republic of China, China National Institute of Standardization, Hubei Exit-Entry Inspection and Quarantine Bureau of the People's Republic of China,
The main drafters of this standard are Wang Douwen, Chen Wei, San Zhenfan, Chen Chang, Mou Luming, and this standard is an industry standard for exit-entry inspection and quarantine issued by the People's Republic of China. SN/T [492—2034
There are inherent uncertainties in all test methods. When testing materials, the method is expected to obtain test results equivalent to the acceptance standard (ARV). If the test result is close to the ARV value, it can be guaranteed that the test method performed by the laboratory has no deviation, or even if there is a deviation, it can be considered negligible and not considered: If the difference between the test result and the ARV exceeds the specified amount, the laboratory must take corresponding improvement measures: For laboratories that can adopt the standard, when testing CS or CCS materials, they can establish acceptable tolerances based on the combined or -uncertainty (e) and the expected first-class error to ensure that the implementation of the test method is not deviated. This standard uses the set tolerance interval to correctly evaluate the detection efficacy of different deviations and examine the detection capability range of true deviations. And further, according to the required first-class error and deviation detection capabilities, to meet the specific needs of the industry. The CS test described in this standard is a supplement to the internal quality control method of the laboratory space - MR statistical combined dynamic tracking and monitoring technology. This standard assumes that the measured system is under normal stress and can be described and predicted using a normal (Gaussian) distribution. Range
Method for evaluating laboratory deviation using a single test result of a standard material
SN/T 1492—2004
This standard specifies methods for evaluating the difference between a test result of a standard (CS) material and its standard value (AR) without a defined and acceptable tolerance interval, and for calculating the probability that a single test result falls outside the tolerance interval. This standard is applicable to the determination of whether a laboratory test method that produces consistent numerical results has deviations, and is not applicable to the evaluation of a single test result obtained by taking a CS material in a certain period of time under field conditions of density or repeatability. 2 Normative references
The clauses in the following documents become the clauses of the standard through reference in this standard. For dated references, all subsequent amendments (excluding errata) or revisions do not apply to this standard. However, parties to agreements based on this standard are encouraged to investigate whether the latest versions of these documents can be used. For undated references, the latest versions shall apply to this standard. B/T5-187 Determination of octane number of automobiles (research method) ASTMIJ6299 Method for evaluating and analyzing the performance of measurement systems using statistical quality assurance techniques ASTME178 Method for handling out-of-bounds observations
3 Terms and definitions
The following terms and definitions apply to this standard. 3.1
Acceptable tolerance interval acccptabl:tolcranczonc is limited to 0_e (the value is determined by the user's selected error. E value see 3.7): If the difference between the results of the implementation of the test method and its ARV falls within the range, the laboratory or test organization shall accept the assumption that the test method operation has not deviated from the assumptions made, and the difference can be regarded as the contribution of the normal random fluctuation of the test method. On the contrary, if the difference falls outside the interval, the assumptions made are not accepted.
Consistency verification standard consensus Check standard (CCS) A typical CS material, HAVR is assigned by collecting at least 16 out-of-bounds (ASTM E178) test results under reproducibility conditions: the results are passed by the Anlcrson-Irling (A) normality test of AS1M I299 standard. I95 is a small confidence level for other normal statistical tests to calculate the arithmetic mean. Note: CS can be a product with insufficient components or a material with special components that can be reproduced, but the former has a test method and often tests the composition of the material, while the latter does not.
Deviation detection power or bias detection When using a standard method, it means that the least deviation can be accurately detected: under given slow conditions and within the tolerance interval. The deviation detection power is defined as "I-II type error". Note: "I-II type error" is often called detection efficiency in traditional general hypothesis testing, and it is the probability of correcting the hypothesis when the mark and alternative are set as points. In the application of this standard, the detection efficiency is the rate of at least positive and negative deviations. SN/T [492—2034
Standardized A(As) standardized delta(As)△ is expressed in units of uncertainty (e) as shown in equation (1): (A.) = /s
Standard error of accepted standard value (SEarv)Standard error of ARv (SEary)The uncertainty of the ARV value is a quantitative measure of the uncertainty of the ARV, which is used as an estimate of the value of the relevant characteristic. For CCS, it is defined as: CocaVN
Wu:
-The total number of out-of-range results collected under reproducible conditions used to confirm the ARV value; N
Standard deviation of all out-of-range results.
Note: Assume a normal model. The 35% confidence margin of the value of the relevant characteristic is (3): ARV-1.96SER~ANV 1.96SEAR
Combined uncertainty () totalumccrtaimty(t) is the combined quantity of the test method stTE and SEARv, expressed as follows: t - V(orE 1 SFARV)
Type T error type T error
In the method proposed in this standard, it refers to the theoretical long-term probability that the correct hypothesis made by the test method without deviation is rejected, resulting in a wrong decision.
Jiang: In the traditional statistical hypothesis test, the first type of error is usually called alpha error: This error refers to the statistical probability of making a wrong assumption or null hypothesis based on the relevant data. The method proposed in this standard regards the null hypothesis as: the test method is implemented without deviation; or it can also be regarded as I, which is different from the C expression. bzxz.net
Type II error type II error
In this method, it refers to the long-term probability that the false hypothesis caused by the deviation of the test method is accepted (i.e., rejected) when the operation of the test method has at least △ deviations, which leads to the long-term probability of making a wrong decision. Note: In traditional statistical tests, type II errors are often called β errors. This error refers to the probability of making this error when the null hypothesis is incorrectly accepted due to the statistical analysis of relevant data. In order to quantify type II errors, it is required to use and propose a certain alternative hypothesis. The alternative hypothesis in the method is: "the test method deviates by a certain amount". The deviation can be considered by the user in advance as the minimum deviation (upper or lower deviation) that is actually reasonable. It can also be considered that the hypothesis is not, that is, III:1 is expressed as follows 4 operation steps
4.1 Determine the suitability of CS by evaluating the [SE/as] ratio Note: Less than or equal to 0. = is considered suitable.
4.2 Calculate the required error rate (GF 1 SEAu).
4.3 Specify the required type II error rate. Note: The recommended starting point is 0.05.
4.4 Specify the required A quantity,
Note: The specification of the quantity value is usually due to commercial risks or operational issues, or the requirements of the two old bands, [statistical considerations. 4.5 Calculate the A-2/e.
4.6 vibration calculated △ 5 and the required type 1 error, check the corresponding test power (see Table 1) 2
Type I error rate
1 Type I error and related deviation test power under different A5 values Table 1
Confirm test power (area)
If the test power is not appropriate
Satisfies:
Type I error;
Deviation test power.
0. 640.73
Heart,69
0.930, 98
Heart, 9
SN/T1492—2004
Generally, if the value is too low, the following items can be selected one or all in turn until all requirements are met. Note: For a single test result of the test method 1, the deviation detection number is determined according to the color quantity, uncertainty of the combined effect and the first error rate selected. This function is sufficient to increase with the increase of the first error rate or the market. According to the first error selected in the test method 1 and the deviation detection function, the appropriate value is used to calculate the acceptance tolerance zone. 4.8
4.9 Establish the acceptance tolerance interval: kE~kE. 4. 1CCS material single test result X is accepted, the result is determined by the ARV value. X-ARV falls within the acceptance tolerance range, and the laboratory implements the test without deviation. X-ARV falls outside the acceptance tolerance range in a positive bias position, and the laboratory implements the test method with a positive deviation of the center value. X-ARV falls outside the acceptance tolerance range in a negative bias position, and the laboratory implements the test method with at least A negative deviation. For examples, see Appendix A. SN/T [492-2034] A.1 The following conditions must be met when applying this standard Appendix A (Normative Appendix) A, 1.1 Standard materials should have their ARV values and their standard deviations (SEar), and for a certain test function, the relevant deviations and 5-[ Since the ratio of SER to SE is 0.5, efforts should be made to ensure that the actual ratio of SER/CSIE is as low as possible. A ratio of 0.5 or less than C.5 is considered appropriate.
A.1.2 The user has a reasonable value for the test method: This value should be reasonable for the standard material: In some cases, CS may be different from the material or components of the commonly tested material, the performance level may not be close, or both situations may exist. In this case, the commonly tested material with the closest performance level (S) should be used to estimate the standard deviation (s) of the field precision. A.1.3 The user should specify the required type I error and the minimum deviation value △ that is actually considered. A.1.4 The test method is under statistical control, that is, it has stability and is passive to accidental causes, but it is allowed to deviate from 4.2 Purpose
A laboratory implements GB/T A.3 CCS material situation
A.3.1 There are 3C experimental space sample exchange tests. For the gasoline products issued, each experiment can only be used once. A.3.2 The standard deviation of the collected non-outlier results is C.25. A.3.3 The AD normality test at the confidence level of 95% showed no significant difference. A.4 Operation steps
A.4.1 Estimation of SE, Rv/nsrm ratio to determine the suitability of CS A4.1.1 The gasoline product with a research alkane number of 91.【 was prepared and internal quality control tests were carried out periodically. The estimated value of s!Tr of the B/T5287 test method was 0.1. A, 4.1.2 SEA of CCS material = 0.25/(30)1 = 0, C46.A, 4.1.3 [SLAk/:5sm:] = 0.046/0.1 = 0, 1620.5; The conclusion is that the CS material is suitable for monthly evaluation letter deviation, A, 4.2 Calculation
Calculation method see 4.2.
e-_0.13 1 0.04537/3 = 0.11A.4.3 stipulates the required type I error rate
This requirement recommends that the error rate be determined at 0.05. A.4.4 The required quantity A is determined based on the laboratory's business needs. A is estimated to be between 2 and 0.22. A.4.5 Calculate As. The calculation method is shown in 4.5.
45=△/E-2
A, 4.6 Check the test efficacy. According to the calculated △ value and the first-class error, the test efficacy is . Corrected test efficacy. The laboratory considers that the test efficacy is too low, so it searches for the corresponding values according to the first-class error in the test. Finally, it is considered that 0.2 (20%) is appropriate, and the corresponding high-definition test efficacy of its △ is determined to be 9.76 or 76%. A. 4. 8 Check k value
According to the type 1 error and the probability of detection selected in Table 1, we get .28.
A4.9 Calculate the acceptance tolerance interval
Calculation method see 1.9. The acceptance tolerance interval is 0±1.28×0.11. That is: -0.14~10.14A.4.10 Calculate (X-ARV) quantity
The experimental study (CS material received a single result of 92.5, then (X92.2)-92.592.2-0.3
A.5 Judgment
A.5.1G.3 falls outside the acceptance tolerance area, so the GB/T5487 test method implemented by the laboratory has deviations (see Chapter 5) A.5.2 The experimental firm implemented (G13/T5187 test square deviation of at least C.3 see Chapter 5). SN/T 1492-2004
Entry and Border Inspection and Quarantine of the People's Republic of China
Industry Standards
Method for evaluating laboratory deviation using single test results of standard materials
SN/T1492—2004
Published by China Standards Publishing House
No. 1, Hebei Street, Sanxian County, Yafuxingmen Office
Postal code: 00045
Website: hvohs.eom
Tel: 58523915
68517348
Printed by China Standards Publishing House, Huangdao Printing Factory, 880×1230 =/16 Sheet 0.75 Word count 1=1000 2C05 February first edition 2C5 February first pre-press*
Book 5: 155063·2-16=017 Price 8.00 yuan If there is any printing error, the distribution center of our company will replace it. Copyright infringement will be investigated
Report phone: (010) 68533533Value: This value should be reasonable for standard materials: In some cases, CS may be similar to the materials or components of the commonly tested materials, the performance level is not close, or the two situations do not exist. In this case, the commonly tested material with the closest performance level (S) should be used to estimate the standard deviation (s) of the field precision. A.1.3 The user should specify the required type I error and the minimum deviation value △ for practical consideration. A.1.4 The test method is under statistical control, that is, it has stability and is passive under accidental causes, but it is allowed to have deviations. 4.2 Purpose
A laboratory implements the GB/T 5187 method and uses CCS material to measure a single result of 92.5. When the ARV value is $2.2, it is hoped to evaluate the implementation of GB/T Is there any deviation in the 5287 test method? A.3 CCS material situation
A.3.1 There are 3C experimental chambers for space sample exchange tests. For the gasoline products issued, each experiment can only be used once. A.3.2 The standard deviation of the non-out-of-bounds results collected is C.25. A.3.3 There is no significant difference in the AD normal test at the confidence level of 95. A, 4 Operation steps
A, 4. 1 Estimate the ratio of SE, Rv/nsrm] to determine the suitability of CS A4.1.1 The gasoline preparation product with a research method alkyl number of 91.【 was used for internal quality control testing. The estimated value of s!Tr of the B/T5287 test method was 0.1. A, 4. 1.2 The SEA of CCS material is 0. 25/(30)1=0,C46.A,4.1.3[SLAk/:5sm:]=0.046/0.1=0,1620.5; The conclusion is that the CS material is suitable for monthly evaluation letter deviation,A, 4.2 Calculation
Calculation method see 4.2.
e-_0. 13 1 0. 04537/3 =0. 11A.4.3 stipulates the required type I error rate
This requirement recommends that the error rate be determined at 0. 05. A.4.4 The required 4 quantities are specified based on the laboratory's business needs. A is estimated to be between 2 and 0.22. A.4.5 Calculate As. The calculation method is shown in 4.5.
45=△/E-2
A, 4.6 Check the test efficacy. According to the calculated △ value and the first-class error, the test efficacy is . Corrected test efficacy. The laboratory considers that the test efficacy is too low, so it searches for the corresponding values according to the first-class error in the error. Finally, it is considered that 0.2 (20%) is appropriate, and the corresponding high-confirmation test efficacy of its △ quantity is determined to be 9.76 or 76%. A. 4. 8 Check k value
According to the type 1 error and probability of detection selected in Table 1, we get .28.
A4.9 Calculate the acceptance tolerance interval
Calculation method see 1.9. The acceptance tolerance interval is 0±1.28×0.11. That is: -0.14~10.14A.4.10 Calculate (X-ARV) quantity
The experimental study (CS material received a single result of 92.5, then (X92.2)-92.592.2-0.3
A.5 Judgment
A.5.1G.3 falls outside the acceptance tolerance area, so the GB/T5487 test method implemented by this laboratory has deviations (see Chapter 5) A.5.2 The experimental implementation (G13/T5187 test square deviation is at least C.3 see Chapter 5). SN/T 1492-2004
Entry/Exit Inspection and Quarantine of the People's Republic of China
Industry Standards
Method for evaluating laboratory deviation using single test results of standard materials
SN/T1492—2004
Published by China Standards Publishing House
No. 1, Hebei Street, Sanxian County, Yafuxingmen Office
Postal code: 00045
Website: hvohs.eom
Tel: 58523915
68517348
Printed by China Standards Publishing House, Huangdao Printing Factory, 880×1230 =/16 Sheet 0.75 Word count 1=1000 2C05 February first edition 2C5 February first pre-press*
Book 5: 155063·2-16=017 Price 8.00 yuan If there is any printing error, the distribution center of our company will replace it. Copyright infringement will be investigated
Report phone: (010) 68533533Value: This value should be reasonable for standard materials: In some cases, CS may be similar to the materials or components of the commonly tested materials, the performance level is not close, or the two situations do not exist. In this case, the commonly tested material with the closest performance level (S) should be used to estimate the standard deviation (s) of the field precision. A.1.3 The user should specify the required type I error and the minimum deviation value △ for practical consideration. A.1.4 The test method is under statistical control, that is, it has stability and is passive under accidental causes, but it is allowed to have deviations. 4.2 Purpose
A laboratory implements the GB/T 5187 method and uses CCS material to measure a single result of 92.5. When the ARV value is $2.2, it is hoped to evaluate the implementation of GB/T Is there any deviation in the 5287 test method? A.3 CCS material situation
A.3.1 There are 3C experimental chambers for space sample exchange tests. For the gasoline products issued, each experiment can only be used once. A.3.2 The standard deviation of the non-out-of-bounds results collected is C.25. A.3.3 There is no significant difference in the AD normal test at the confidence level of 95. A, 4 Operation steps
A, 4. 1 Estimate the ratio of SE, Rv/nsrm] to determine the suitability of CS A4.1.1 The gasoline preparation product with a research method alkyl number of 91.【 was used for internal quality control testing. The estimated value of s!Tr of the B/T5287 test method was 0.1. A, 4. 1.2 The SEA of CCS material is 0. 25/(30)1=0,C46.A,4.1.3[SLAk/:5sm:]=0.046/0.1=0,1620.5; The conclusion is that the CS material is suitable for monthly evaluation letter deviation,A, 4.2 Calculation
Calculation method see 4.2.
e-_0. 13 1 0. 04537/3 =0. 11A.4.3 stipulates the required type I error rate
This requirement recommends that the error rate be determined at 0. 05. A.4.4 The required 4 quantities are specified based on the laboratory's business needs. A is estimated to be between 2 and 0.22. A.4.5 Calculate As. The calculation method is shown in 4.5.
45=△/E-2
A, 4.6 Check the test efficacy. According to the calculated △ value and the first-class error, the test efficacy is . Corrected test efficacy. The laboratory considers that the test efficacy is too low, so it searches for the corresponding values according to the first-class error in the error. Finally, it is considered that 0.2 (20%) is appropriate, and the corresponding high-confirmation test efficacy of its △ quantity is determined to be 9.76 or 76%. A. 4. 8 Check k value
According to the type 1 error and probability of detection selected in Table 1, we get .28.
A4.9 Calculate the acceptance tolerance interval
Calculation method see 1.9. The acceptance tolerance interval is 0±1.28×0.11. That is: -0.14~10.14A.4.10 Calculate (X-ARV) quantity
The experimental study (CS material received a single result of 92.5, then (X92.2)-92.592.2-0.3
A.5 Judgment
A.5.1G.3 falls outside the acceptance tolerance area, so the GB/T5487 test method implemented by this laboratory has deviations (see Chapter 5) A.5.2 The experimental implementation (G13/T5187 test square deviation is at least C.3 see Chapter 5). SN/T 1492-2004
Entry/Exit Inspection and Quarantine of the People's Republic of China
Industry Standards
Method for evaluating laboratory deviation using single test results of standard materials
SN/T1492—2004
Published by China Standards Publishing House
No. 1, Hebei Street, Sanxian County, Yafuxingmen Office
Postal code: 00045
Website: hvohs.eom
Tel: 58523915
68517348
Printed by China Standards Publishing House, Huangdao Printing Factory, 880×1230 =/16 Sheet 0.75 Word count 1=1000 2C05 February first edition 2C5 February first pre-press*
Book 5: 155063·2-16=017 Price 8.00 yuan If there is any printing error, the distribution center of our company will replace it. Copyright infringement will be investigated
Report phone: (010) 6853353300 yuan If there is any printing error, the distribution center of our company will replace it. Copyright infringement will be investigated.
Report phone: (010)6853353300 yuan If there is any printing error, the distribution center of our company will replace it. Copyright infringement will be investigated.
Report phone: (010)68533533
Tip: This standard content only shows part of the intercepted content of the complete standard. If you need the complete standard, please go to the top to download the complete standard document for free.