Petroleum products - Determination and application of precision data in relation to methods of test

ISO 4259:2006 covers the calculation of precision estimates and their application to specifications. In particular, it contains definitions of relevant statistical terms, the procedures to be adopted in the planning of an inter-laboratory test programme to determine the precision of a test method, the method of calculating the precision from the results of such a programme, and the procedure to be followed in the interpretation of laboratory results in relation both to precision of the test methods and to the limits laid down in specifications. The procedures in ISO 4259:2006 have been designed specifically for petroleum and petroleum-related products, which are normally homogeneous. However, the procedures described in this International Standard can also be applied to other types of homogeneous products. Careful investigations are necessary before applying ISO 4259:2006 to products for which the assumption of homogeneity can be questioned.

Produits pétroliers — Détermination et application des valeurs de fidélité relatives aux méthodes d'essai

L'ISO 4259:2006 traite du calcul des estimations de fidélité et de leur application aux spécifications. En particulier, elle contient les définitions des termes statistiques concernés, les procédures à suivre dans l'organisation d'un programme d'essai interlaboratoires destiné à déterminer la fidélité d'une méthode d'essai, la méthode de calcul de la fidélité à partir des résultats d'un tel programme et la procédure à suivre dans l'interprétation des résultats de laboratoire, à la lumière de la fidélité des méthodes de test et des limites fixées dans les spécifications. Les procédures de l'ISO 4259:2006 ont été conçues spécifiquement pour les produits pétroliers et leurs produits connexes, qui sont normalement homogènes. Cependant on reconnaît que les procédures décrites dans l'ISO 4259:2006 peuvent aussi s'appliquer à d'autres types de produits homogènes. Des contrôles attentifs s'avèrent nécessaires avant d'appliquer l'ISO 4259:2006 à des produits pour lesquels la présomption d'homogénéité peut être mise en question.

General Information

Status
Withdrawn
Publication Date
27-Jul-2006
Withdrawal Date
27-Jul-2006
Current Stage
9599 - Withdrawal of International Standard
Start Date
01-Nov-2017
Completion Date
13-Dec-2025
Ref Project

Relations

Effective Date
28-Feb-2023
Standard
ISO 4259:2006 - Petroleum products -- Determination and application of precision data in relation to methods of test
English language
62 pages
sale 15% off
Preview
sale 15% off
Preview
Standard
ISO 4259:2006 - Produits pétroliers -- Détermination et application des valeurs de fidélité relatives aux méthodes d'essai
French language
64 pages
sale 15% off
Preview
sale 15% off
Preview
Standard
ISO 4259:2006
Russian language
12 pages
sale 15% off
Preview
sale 15% off
Preview

Frequently Asked Questions

ISO 4259:2006 is a standard published by the International Organization for Standardization (ISO). Its full title is "Petroleum products - Determination and application of precision data in relation to methods of test". This standard covers: ISO 4259:2006 covers the calculation of precision estimates and their application to specifications. In particular, it contains definitions of relevant statistical terms, the procedures to be adopted in the planning of an inter-laboratory test programme to determine the precision of a test method, the method of calculating the precision from the results of such a programme, and the procedure to be followed in the interpretation of laboratory results in relation both to precision of the test methods and to the limits laid down in specifications. The procedures in ISO 4259:2006 have been designed specifically for petroleum and petroleum-related products, which are normally homogeneous. However, the procedures described in this International Standard can also be applied to other types of homogeneous products. Careful investigations are necessary before applying ISO 4259:2006 to products for which the assumption of homogeneity can be questioned.

ISO 4259:2006 covers the calculation of precision estimates and their application to specifications. In particular, it contains definitions of relevant statistical terms, the procedures to be adopted in the planning of an inter-laboratory test programme to determine the precision of a test method, the method of calculating the precision from the results of such a programme, and the procedure to be followed in the interpretation of laboratory results in relation both to precision of the test methods and to the limits laid down in specifications. The procedures in ISO 4259:2006 have been designed specifically for petroleum and petroleum-related products, which are normally homogeneous. However, the procedures described in this International Standard can also be applied to other types of homogeneous products. Careful investigations are necessary before applying ISO 4259:2006 to products for which the assumption of homogeneity can be questioned.

ISO 4259:2006 is classified under the following ICS (International Classification for Standards) categories: 75.080 - Petroleum products in general. The ICS classification helps identify the subject area and facilitates finding related standards.

ISO 4259:2006 has the following relationships with other standards: It is inter standard links to ISO/R 1161:1970, ISO 4210-2:2014, ISO 4259-2:2017, ISO 4259-1:2017, SIST ISO 4259:1996, ISO 4259:1992/Cor 1:1993, ISO 4259:1992. Understanding these relationships helps ensure you are using the most current and applicable version of the standard.

You can purchase ISO 4259:2006 directly from iTeh Standards. The document is available in PDF format and is delivered instantly after payment. Add the standard to your cart and complete the secure checkout process. iTeh Standards is an authorized distributor of ISO standards.

Standards Content (Sample)


INTERNATIONAL ISO
STANDARD 4259
Third edition
2006-08-01
Petroleum products — Determination and
application of precision data in relation to
methods of test
Produits pétroliers — Détermination et application des valeurs de
fidélité relatives aux méthodes d'essai

Reference number
©
ISO 2006
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.

©  ISO 2006
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland
ii © ISO 2006 – All rights reserved

Contents Page
Foreword. v
Introduction . vi
1 Scope . 1
2 Normative references . 1
3 Terms and definitions. 1
4 Stages in the planning of an inter-laboratory test programme for the determination of the
precision of a test method . 4
4.1 General. 4
4.2 Preparing a draft method of test . 5
4.3 Planning a pilot programme with at least two laboratories. 5
4.4 Planning the inter-laboratory programme. 5
4.5 Executing the inter-laboratory programme.5
5 Inspection of inter-laboratory results for uniformity and for outliers . 6
5.1 General. 6
5.2 Transformation of data. 7
5.3 Tests for outliers. 8
5.4 Rejection of complete data from a sample . 11
5.5 Estimating missing or rejected values . 12
5.6 Rejection test for outlying laboratories. 13
5.7 Confirmation of selected transformation .14
6 Analysis of variance, calculation and expression of precision estimates . 14
6.1 General. 14
6.2 Analysis of variance . 14
6.3 Expectation of mean squares and calculation of precision estimates . 17
6.4 Expression of precision estimates of a method of test. 20
7 Significance of repeatability (r) and reproducibility (R). 21
7.1 General. 21
7.2 Repeatability, r . 21
7.3 Reproducibility, R.22
8 Specifications. 24
8.1 Aim of specifications. 24
8.2 Construction of specifications limits in relation to precision. 24
9 Quality control against specifications. 25
9.1 General. 25
9.2 Testing margin at the supplier . 25
9.3 Testing margin at the recipient . 25
10 Dispute procedure . 25
Annex A (normative) Determination of number of samples required. 28
Annex B (informative) Derivation of equation for calculating the number of samples required . 30
Annex C (normative) Notation and tests. 31
Annex D (normative) Example results of test for determination of bromine number and statistical
tables. 36
Annex E (normative) Types of dependence and corresponding transformations. 45
Annex F (normative) Weighted linear regression analysis. 49
Annex G (normative) Rules for rounding off results . 56
Annex H (informative) Explanation of equations given in Clause 7 . 57
Annex I (informative) Specifications that relate to a specified degree of criticality. 59
Bibliography . 62

iv © ISO 2006 – All rights reserved

Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 4259 was prepared by Technical Committee ISO/TC 28, Petroleum products and lubricants.
This third edition cancels and replaces the second edition (ISO 4259:1992), Clauses 1, 5, 7 C.7, E.2 and F.3
and subclauses 4.2, 5.2, 6.3.2, 6.3.3.1, 6.3.3.3, 6.4, 8.2, 10.2, 10.4 and 10.5, which have been technically
revised. It also incorporates the Technical Corrigendum ISO 4259:1992/Cor.1:1993.

Introduction
For purposes of quality control and to check compliance with specifications, the properties of commercial
petroleum products are assessed by standard laboratory test methods. Two or more measurements of the
same property of a specific sample by any given test method do not usually give exactly the same result. It is,
therefore, necessary to take proper account of this fact, by arriving at statistically-based estimates of the
precision for a method, i.e. an objective measure of the degree of agreement expected between two or more
results obtained in specified circumstances.
[11]
ISO 4259 makes reference to ISO 3534-2 , which gives a different definition of true value (see 3.26).
ISO 4259 also refers to ISO 5725-2. The latter is required in particular and unusual circumstances (see 5.2)
for the purpose of estimating precision.

vi © ISO 2006 – All rights reserved

INTERNATIONAL STANDARD ISO 4259:2006(E)

Petroleum products — Determination and application of
precision data in relation to methods of test
1 Scope
This International Standard covers the calculation of precision estimates and their application to specifications.
In particular, it contains definitions of relevant statistical terms (Clause 3), the procedures to be adopted in the
planning of an inter-laboratory test programme to determine the precision of a test method (Clause 4), the
method of calculating the precision from the results of such a programme (Clauses 5 and 6), and the
procedure to be followed in the interpretation of laboratory results in relation both to precision of the test
methods and to the limits laid down in specifications (Clauses 7 to 10).
The procedures in this International Standard have been designed specifically for petroleum and petroleum-
related products, which are normally homogeneous. However, the procedures described in this International
Standard can also be applied to other types of homogeneous products. Careful investigations are necessary
before applying this International Standard to products for which the assumption of homogeneity can be
questioned.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO 5725-2:1994, Accuracy (trueness and precision) of measurement methods and results — Part 2: Basic
method for the determination of repeatability and reproducibility of a standard measurement method
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
3.1
analysis of variance
technique that enables the total variance of a method to be broken down into its component factors
3.2
between-laboratory variance
element of the total variance attributable to the difference between the mean values of different laboratories
NOTE 1 When results obtained by more than one laboratory are compared, the scatter is usually wider than when the
same number of tests are carried out by a single laboratory, and there is some variation between means obtained by
different laboratories. These give rise to the between-laboratory variance which is that component of the overall variance
due to the difference in the mean values obtained by different laboratories.
NOTE 2 There is a corresponding definition for between-operator variance.
NOTE 3 The term “between-laboratory” is often shortened to “laboratory” when used to qualify representative
parameters of the dispersion of the population of results, for example as “laboratory variance”.
3.3
bias
difference between the true value (related to the method of test) and the known value, where this is available
NOTE For a definition of “true value” and “known value,” see 3.26 and 3.8, respectively.
3.4
blind coding
assignment of a different number to each sample so that no other identification or information on any sample
is given to the operator
3.5
check sample
sample taken at the place where the product is exchanged, i.e. where the responsibility for the product quality
passes from the supplier to the recipient
3.6
degrees of freedom
divisor used in the calculation of variance; one less than the number of independent results
NOTE The definition applies strictly only in the simplest cases. Complete definitions are beyond the scope of this
International Standard.
3.7
determination
process of carrying out the series of operations specified in the test method, whereby a single value is
obtained
3.8
known value
actual quantitative value implied by the preparation of the sample
NOTE The known value does not always exist, for example for empirical tests such as flash point.
3.9
mean
arithmetic mean
sum of the results divided by their number for a given set of results
3.10
mean square
sum of squares divided by the degrees of freedom
3.11
normal distribution
probability distribution of a continuous random variable, x, such that, if x is any real number, the probability
density is
⎡⎤
11 x − µ
⎛⎞
⎢⎥
fx=−exp ,−∞ ()
⎜⎟
2 σ
σ 2π⎢⎥⎝⎠
⎣⎦
NOTE µ is the true value and σ is the standard deviation of the normal distribution (σ > 0).
3.12
operator
person who normally and regularly carries out a particular test
2 © ISO 2006 – All rights reserved

3.13
outlier
result far enough in magnitude from other results to be considered not a part of the set
3.14
precision
closeness of agreement between the results obtained by applying the experimental procedure several times
on identical materials and under prescribed conditions
NOTE The smaller the random part of the experimental error, the more precise is the procedure.
3.15
random error
chance variation encountered in all test work despite the closest control of variables
3.16
recipient
any individual or organization who receives or accepts the product delivered by the supplier
3.17
repeatability
〈qualitatively〉 closeness of agreement between independent results obtained in the normal and correct
operation of the same method on identical test material, in a short interval of time, and under the same test
conditions (same operator, same apparatus, same laboratory)
NOTE The representative parameters of the dispersion of the population that can be associated with the results are
qualified by the term “repeatability”, for example, repeatability standard deviation or repeatability variance. It is important
that the term “repeatability” not be confused with the terms “between repeats” or “repeats” when used in this way (see
3.19). Repeatability refers to the state of minimum random variability of results. The period of time during which repeated
results are to be obtained should therefore be short enough to exclude time-dependent errors, for example, environmental
and calibration errors.
3.18
repeatability
〈quantitatively〉 value equal to or below which the absolute difference between two single test results obtained
in the conditions specified that can be expected to lie with a probability of 95 %
NOTE For the details of the conditions specified, see 3.17.
3.19
replication
execution of a test method more than once so as to improve precision and to obtain a better estimation of
testing error
NOTE Replication should be distinguished from repetition in that the former implies that repeated experiments are
carried out at one place and, as far as possible, within one period of time. The representative parameters of the dispersion
of the population that can be associated with repeated experiments are qualified by the term “between repeats”, or in
shortened form “repeats”, for example, “repeats standard deviation”.
3.20
reproducibility
〈qualitatively〉 closeness of agreement between individual results obtained in the normal and correct operation
of the same method on identical test material but under different test conditions (different operators, different
apparatus and different laboratories)
NOTE The representative parameters of the dispersion of the population that can be associated with the results are
qualified by the term “reproducibility”, for example, reproducibility standard deviation or reproducibility variance.
3.21
reproducibility
〈quantitatively〉 value equal to or below which the absolute difference between two single test results on
identical material obtained by operators in different laboratories, using the standardized test method, may be
expected to lie with a probability of 95 %
3.22
result
final value obtained by following the complete set of instructions in the test method; it may be obtained from a
single determination or from several determinations depending on the instructions in the method
NOTE It is assumed that the result is rounded off according to the procedure specified in Annex G.
3.23
standard deviation
measure of the dispersion of a series of results around their mean, equal to the positive square root of the
variance and estimated by the positive square root of the mean square
3.24
sum of squares
sum of squares of the differences between a series of results and their mean
3.25
supplier
any individual or organization responsible for the quality of a product just before it is taken over by the
recipient
3.26
true value
for practical purposes, the value towards which the average of single results obtained by n laboratories tends,
as n tends towards infinity
NOTE 1 Such a true value is associated with the particular method of test.
[11]
NOTE 2 A different and idealized definition is given in ISO 3534-2 .
3.27
variance
mean of the squares of the deviation of a random variable from its mean, estimated by the mean square
4 Stages in the planning of an inter-laboratory test programme for the
determination of the precision of a test method
4.1 General
The stages in planning an inter-laboratory test programme are as follows:
a) preparing a draft method of test;
b) planning a pilot programme with at least two laboratories;
c) planning the inter-laboratory programme;
d) executing the inter-laboratory programme.
The four stages are described in turn in 4.2 to 4.5.
4 © ISO 2006 – All rights reserved

4.2 Preparing a draft method of test
This shall contain all the necessary details for carrying out the test and reporting the results. Any condition that
could alter the results shall be specified.
A clause on precision is included in the draft method of the test at this stage only as a heading. It is
recommended that the lower limit of the scope of the test method is not less than the region of the lowest
value tested in the inter-laboratory programme, and is at least 2R greater than the lowest achievable result
(see 8.2), where R is the reproducibility estimate. Similarly, it is recommended that the upper limit of the scope
of a test method is not greater than the region of the highest value tested in the inter-laboratory programme,
and is at least 2R less than the highest achievable result.
4.3 Planning a pilot programme with at least two laboratories
A pilot programme is necessary for the following reasons:
a) to verify the details in the operation of the test;
b) to find out how well operators can follow the instructions of the method;
c) to check the precautions regarding samples;
d) to estimate approximately the precision of the test.
At least two samples are required, covering the range of results to which the test method is intended to apply;
however, at least twelve laboratory/sample combinations shall be included. Each sample is tested twice by
each laboratory under repeatability conditions. If any omissions or inaccuracies in the draft test method are
revealed, they shall now be corrected. The results shall be analysed for bias and precision; if either is
considered to be too large, then alterations to the test method shall be considered.
4.4 Planning the inter-laboratory programme
There shall be at least five participating laboratories, but it is preferable that there are more in order to reduce
the number of samples required.
The number of samples shall be sufficient to cover the range of the property measured at approximately
equidistant intervals and to give reliability to the precision estimates. If precision is found to vary with the level
of results in the pilot programme, then at least five samples shall be used in the inter-laboratory programme. In
any case, it is necessary to obtain at least 30 degrees of freedom in both repeatability and reproducibility. For
repeatability, this means obtaining a total of at least 30 pairs of results in the programme.
For reproducibility, Table A.1 gives the minimum number of samples required in terms of L, P and Q, where L
is the number of participating laboratories, and P and Q are the ratios of variance component estimates
obtained from the pilot programme. Specifically, P is the ratio of the interaction component to the repeats
component and Q is the ratio of the laboratories component to the repeats component. Annex B gives the
derivation of the equation used. If Q is much larger than P, then 30 degrees of freedom cannot be achieved;
the blank entries in Table A.1 correspond to, or an approach to, this situation (i.e. when more than 20 samples
are required). For these cases, there is likely to be a significant bias between laboratories.
4.5 Executing the inter-laboratory programme
One person shall be responsible for the entire programme, from the distribution of the texts of the test method
and samples to the final appraisal of the results. He shall be familiar with the test method, but shall not
personally take part in the tests.
The text of the test method shall be distributed to all the laboratories in time to allow any queries to be raised
before the tests begin. If any laboratory wants to practice the method in advance, this shall be carried out with
samples other than those used in the programme.
The samples shall be accumulated, subdivided and distributed by the organizer, who shall also keep a reserve
of each sample for emergencies. It is most important that the individual laboratory portions be homogeneous.
They shall be blind coded before distribution and the following information shall be sent with them:
a) agreed (draft) method of test;
b) handling and storage requirements for the samples;
c) order in which the samples are to be tested (a different random order for each laboratory);
d) statement that two results shall be obtained consecutively on each sample by the same operator with the
same apparatus. For statistical reasons, it is imperative that the two results are obtained independently of
each other, that is, that the second result is not biased by knowledge of the first. If this is regarded as
impossible to achieve with the operator concerned, then the pairs of results shall be obtained in a blind
fashion, but ensuring that they are carried out in a short period of time;
e) period of time during which repeated results are to be obtained and the period of time during which all the
samples are to be tested;
f) blank form for reporting the results. For each sample, there shall be space for the date of testing, the two
results, and any unusual occurrences. The unit of accuracy for reporting the results shall be specified;
g) statement that the test shall be carried out under normal conditions, using operators with good experience
but not exceptional knowledge and that the duration of the test shall be the same as normal.
The pilot-programme operators may take part in the inter-laboratory programme. If their extra experience in
testing a few more samples produces a noticeable effect, it serves as a warning that the test method is not
satisfactory. They shall be identified in the report of the results so that any effect can be noted.
5 Inspection of inter-laboratory results for uniformity and for outliers
5.1 General
In 5.2 to 5.7, procedures are specified for examining the results reported in a statistically designed inter-
laboratory programme (see Clause 4) in order to establish the following:
a) independence or dependence of precision and the level of results;
b) uniformity of precision from laboratory to laboratory;
c) and to detect the presence of outliers.
The procedures are described in mathematical terms based on the notation of Annex C and illustrated with
reference to the example data (calculation of bromine number) set out in Annex D.
Throughout 5.2 to 5.7 (and Clause 6), the procedures used are first specified and then illustrated by a worked
example using data given in Annex D.
It is assumed throughout this clause that all the results are either from a single normal distribution or capable
of being transformed into such a distribution (see 5.2). Other cases (which are rare) require a different
treatment that is beyond the scope of this International Standard. See Reference [8] for a statistical test on
normality.
Although the procedures shown here are in a form suitable for hand calculation, it is strongly advised that an
electronic computer with appropriately validated software be used to store and analyse inter-laboratory test
results, based on the procedures of this International Standard (see, for example, Reference [9]).
6 © ISO 2006 – All rights reserved

5.2 Transformation of data
5.2.1 General
In many test methods, the precision depends on the level of the test result, and thus the variability of the
reported results is different from sample to sample. The method of analysis outlined in this International
Standard requires that this shall not be so and the position is rectified, if necessary, by a transformation.
The laboratories standard deviations, D , and the repeats standard deviations, d , for sample j (see Annex C)
j j
are calculated and plotted separately against the sample means, m . If the points so plotted can be considered
j
as lying about a pair of lines parallel to the m-axis, then no transformation is necessary. If, however, the
plotted points describe non-horizontal straight lines or curves of the form D = f (m) and d = f (m), then a
1 2
transformation is necessary.
The relationships D = f (m) and d = f (m) are not, in general, identical. The statistical procedures of this
1 2
International Standard require, however, that the same transformation be applicable both for repeatability and
for reproducibility. For this reason, the two relationships are combined into a single dependency relationship
D = f(m) (where D now includes d) by including a dummy variable, T. This takes account of the difference
between the relationships, if one exists, and provides a means of testing for this difference (see Clause F.1).
The single relationship D = f(m) is best estimated by a weighted linear regression analysis, even though in
most cases an unweighted regression gives a satisfactory approximation. The derivation of weights is
described in Clause F.2, and the computational procedure for the regression analysis is described in
Clause F.3. Typical forms of dependence D = f(m) are given in Clause E.1. These are all expressed in terms of
transformation parameters B and B .
The estimation of B and B , and the transformation procedure which follows, are summarized in Clause E.2.
This includes statistical tests for the significance of the regression (i.e. is the relationship D = f(m) parallel to
the m-axis), and for the difference between the repeatability and reproducibility relationships, based at the 5 %
significance level. If such a difference is found to exist, or if no suitable transformation exists, then the
alternative sample-by-sample procedures of ISO 5725-2 shall be used. In such an event, it is not possible to
test for laboratory bias over all samples (see 5.6) or separately estimate the interaction component of variance
(see 6.2).
If it has been shown at the 5 % significance level that there is a significant regression of the form D = f(m),
then the appropriate transformation y = F(x), where x is the reported result, is given by the equation:
dx
Fx =K (2)
()

f x
()
where K is a constant. In that event, all results shall be transformed accordingly and the remainder of the
analysis carried out in terms of the transformed results. Typical transformations are given in Clause E.1.
It is difficult to make the choice of transformation the subject of formalized rules. Qualified statistical
assistance can be required in particular cases. The presence of outliers can affect judgement as to the type of
transformation required, if any (see 5.7).
5.2.2 Worked example
Table 1 lists the values of m, D, and d for the eight samples in the example given in Annex D, correct to three
significant digits. Corresponding degrees of freedom are in parentheses.
Table 1
Sample
3 8 1 4 5 6 2 7
number
m 0,756 1,22 2,15 3,64 10,9 48,2 65,4 114
D 0,066 9 (14) 0,159 (9) 0,729 (8) 0,211 (11) 0,291 (9) 1,50 (9) 2,22 (9) 2,93 (9)
d 0,050 0 (9) 0,057 2 (9) 0,127 (9) 0,116 (9) 0,094 3 (9) 0,527 (9) 0,818 (9) 0,935 (9)

Inspection of the figures in Table 1 shows that both D and d increase with m, the rate of increase diminishing
as m increases. A plot of these figures on log-log paper (i.e. a graph of log D and log d against log m) shows
that the points may reasonably be considered as lying about two straight lines (see Figure F.1) From the
example calculations given in Clause F.4, the gradients of these lines are shown to be the same, with an
estimated value of 0,638. Bearing in mind the errors in this estimated value, the gradient may, for convenience,
be taken as 2/3.
Hence, the same transformation is appropriate both for repeatability and reproducibility, and is given by the
equation:
−23 13
xxd3=x (3)

Since the constant multiplier may be ignored, the transformation thus reduces to that of taking the cube roots
of the reported results (bromine numbers). This yields the transformed data shown in Table D.2, in which the
cube roots are quoted correct to three decimal places.
5.3 Tests for outliers
5.3.1 General
The reported data, or if it has been decided that a transformation is necessary, the transformed results, shall
be inspected for outliers. These are the values that are so different from the remaining data that it can only be
concluded that they have arisen from some fault in the application of the test method or from testing a wrong
sample. Many possible tests may be used and the associated significance levels varied, but those that are
given below have been found to be appropriate for this International Standard. These outlier tests all assume
a normal distribution of errors (see 5.1).
5.3.2 Uniformity of repeatability
5.3.2.1 General
[1]
The first outlier test is concerned with detecting a discordant result in a pair of repeat results. This test
involves calculating the e over all the laboratory/sample combinations. Cochran's criterion at the 1 %
ij
significance level is then used to test the ratio of the largest of these e values over their sum (see
ij
Clause C.5). If its value exceeds the value given in Table D.3, corresponding to one degree of freedom, n
being the number of pairs available for comparison, then the member of the pair farthest from the sample
mean shall be rejected and the process repeated, reducing n by 1, until no more rejections are called for. In
certain cases, this test “snowballs” and leads to an unacceptably large proportion of rejections (say more than
10 %). If this is so, this rejection test shall be abandoned and some or all of the rejected results shall be
retained. An arbitrary decision based on judgement is necessary in this case.
5.3.2.2 Worked example
In the case of the example given in Annex D, the absolute differences (ranges) between transformed repeat
results, i.e. of the pairs of numbers in Table D.2, in units of the third decimal place, are shown in Table 2.
8 © ISO 2006 – All rights reserved

Table 2
Sample
Laboratory
1 2 3 4 5 6 7 8
A 42 21 7 13 7 10 8 0
B 23 12 12 0 7 9 3 0
C 0 6 0 0 7 8 4 0
D 14 6 0 13 0 8 9 32
E 65 4 0 0 14 5 7 28
F 23 20 34 29 20 30 43 0
G 62 4 78 0 0 16 18 56
H 44 20 29 44 0 27 4 32
J 0 59 0 40 0 30 26 0
The largest range is 0,078 for laboratory G on sample 3. The sum of squares of all the ranges is
22 22
0,042++0,021 …+ 0,026+ 0= 0,043 9
0,078
Thus, the ratio to be compared with Cochran's criterion is = 0,138
0,043 9
There are 72 ranges and, as from Table D.3, the criterion for 80 ranges is 0,170 9, this ratio is not significant.
5.3.3 Uniformity of reproducibility
5.3.3.1 General
The following outlier tests are concerned with establishing uniformity in the reproducibility estimate and are
designed to detect either a discordant pair of results from a laboratory on a particular sample or a discordant
[2]
set of results from a laboratory on all samples. For both purposes, the Hawkins' test is appropriate.
This involves forming for each sample, and finally for the overall laboratory averages (see 5.6), the ratio of the
largest absolute deviation of laboratory mean from sample (or overall) mean to the square root of certain sums
of squares (see Clause C.6).
The ratio corresponding to the largest absolute deviation shall be compared with the critical 1 % values given
in Table D.4, where n is the number of laboratory/sample cells in the sample (or the number of overall
laboratory means) concerned and where ν is the degrees of freedom for the sum of squares, which is
additional to that corresponding to the sample in question. In the test for laboratory/sample cells, ν refers to
other samples, but is zero in the test for overall laboratory averages.
If a significant value is encountered for individual samples, the corresponding extreme values shall be omitted
and the process repeated. If any extreme values are found in the laboratory totals, then all the results from
that laboratory shall be rejected.
If the test “snowballs”, leading to an unacceptably large proportion of rejections (say more than 10 %), then
this rejection test shall be abandoned and some or all of the rejected results shall be retained. An arbitrary
decision based on judgement is necessary in this case.
5.3.3.2 Worked example
The application of Hawkins' test to cell means within samples is shown below.
The first step is to calculate the deviations of cell means from respective sample means over the whole array.
These are shown in Table 3, in units of the third decimal place.
The sum of squares of the deviations are then calculated for each sample. These are also shown in Table 3 in
units of the third decimal place.
Table 3
Sample
Laboratory
1 2 3 4 5 6 7 8
A 20 8 14 15 10 48 6 3
B 75 7 20 9 10 47 6 3
C 64 35 3 20 30 4 22 25
D 314 33 18 42 7 39 80 50
E 32 32 30 9 7 18 18 39
F 75 97 31 20 30 8 74 53
G 10 34 32 20 20 61 9 62
H 42 13 4 42 13 21 8 50
J 1 28 22 29 14 8 10 53
Sum of squares 117 15 4 6 3 11 13 17

The cell tested is the one with the most extreme deviation. This was obtained by laboratory D from sample 1.
The appropriate Hawkins' test ratio is therefore
0,314
B*0==,7281
0,117++0,015 .+ 0,017
The critical value, corresponding to n = 9 cells in sample 1 and ν = 56 extra degrees of freedom from the other
samples, is interpolated from Table D.4 as 0,372 9. The test value is greater than the critical value and so the
results from laboratory D on sample 1 are rejected.
As there has been a rejection, the mean value, deviations and sum of squares are recalculated for sample 1,
and the procedure is repeated. The next cell to be tested is that obtained by laboratory F from sample 2. The
Hawkins' test ratio for this cell is:
0,097
B*0==,3542
0,006++0,015 .+ 0,017
The critical value corresponding to n = 9 cells in sample 2 and ν = 55 extra degrees of freedom is interpolated
from Table D.4 as 0,375 6. As the test ratio is less than the critical value, there are no further rejections.
10 © ISO 2006 – All rights reserved

5.4 Rejection of complete data from a sample
5.4.1 General
The laboratories standard deviation and repeats standard deviation shall be examined for any outlying
samples. If a transformation has been carried out or any rejection made, new standard deviations shall be
calculated.
If the standard deviation for any sample is excessively large, it shall be examined with a view to rejecting the
results from that sample.
Cochran's criterion at the 1 % level can be used when the standard deviations are based on the same number
of degrees of freedom. This involves calculating the ratio of the largest of the corresponding sums of squares
(laboratories or repeats, as appropriate) to their total (see Clause C.5). If the ratio exceeds the critical value
given in Table D.3, with n as the number of samples and ν the degrees of freedom, then all the results from
the sample in question shall be rejected. In such an event, care should be taken that the extreme standard
deviation is not due to the application of an inappropriate transformation (see 5.2), or undetected outliers.
There is no optimal test when standard deviations are based on different degrees of freedom. However, the
ratio of the largest variance to that pooled from the remaining samples follows an F-distribution with ν and ν
1 2
degrees of freedom (see Clause C.7). Here ν is the degrees of freedom of the variance in question and ν is
1 2
the degrees of freedom for the remaining samples. If the ratio is greater than the critical value given in
Tables D.6 to D.10, corresponding to a significance level of 0,01/S, where S is the number of samples, then
results from the sample in question shall be rejected.
5.4.2 Worked example
The standard deviations of the transformed results, after the rejection of the pair of results by laboratory D on
sample 1, are given in Table 4 in ascending order of sample mean, correct to three significant digits.
Corresponding degrees of freedom are in parentheses.
Inspection shows that there is no outlying sample amongst these. It is noted that the standard deviations are
now independent of the sample means, which was the purpose of transforming the results.
The figures in Table 5, taken from a test programme on bromine numbers over 100, illustrate the case of a
sample rejection.
It is clear, by inspection, that the laboratories' standard deviation for sample 93 at 15,26 is far greater than the
others. It is noted that the repeats standard deviation in this sample is correspondingly large.
Table 4
Sample number 3 8 1 4 5 6 2 7
Sample mean 0,910 0 1,066 1,240 1,538 2,217 3,639 4,028 4,851
Laboratories standard
0,027 8 (14) 0,047 3 (9) 0,035 4 (13) 0,029 7 (11) 0,019 7 (9) 0,037 8 (9) 0,045 0 (9) 0,041 6 (9)
deviation
Repeats standard
0,021 4 (9) 0,018 2 (9) 0,028 1 (8) 0,016 4 (9) 0,006 3 (9) 0,013 2 (9) 0,016 6 (9) 0,013 0 (9)
deviation
Table 5
Sample number 90 89 93 92 91 94 95 96
Sample mean 96,1 99,8 119,3 125,4 126,0 139,1 139,4 159,5
Laboratories standard deviation 5,10 (8) 4,20 (9) 15,26 (8) 4,40 (11) 4,09 (10) 4,87 (8) 4,74 (9) 3,85 (8)
Repeats standard deviation
1,13 (8) 0,99 (8) 2,97 (8) 0,91 (8) 0,73 (8) 1,32 (8) 1,12 (8) 1,36 (8)
Since laboratory degrees of freedom are not the same over all samples, the variance ratio test is used. The
variance pooled from all samples excluding sample 93 is the sum of the sums of squares divided by the total
degrees of freedom, that is:
22 2
⎡⎤
8×+5,10 9× 4,20+ .+ 8× 3,85
()( ) ( )
⎢⎥
⎣⎦
= 19,96
(8++9 .+ 8)
The variance ratio is then calculated as (15,26 )/19,96 = 11,66.
From Tables D.6 to D.10, the critical value corresponding to a significance level of 0,01/8 = 0,001 25, for 8 and
63 degrees of freedom, is approximately 4. This is less than the test ratio and results from sample 93 shall.
therefore, be rejected.
Turning to repeats standard deviations, it is noted that degrees of freedom are identical for each sample and
that Cochran's test can therefore be applied. Cochran's criterion is the ratio of the largest sum of squares
(sample 93) to the sum of all the sums of squares, that is:
2 2 2 2
2,97 /(1,13 + 0,99 + . + 1,36 ) = 0,510
This is greater than the critical value of 0,352 corresponding to n = 8 and ν = 8 (see Table D.3), and confirms
that results from sample 93 shall be rejected.
5.5 Estimating missing or rejected values
5.5.1 One of the two repeat values missing or rejected
If one of a pair of repeats (y or y ) is missing or rejected, this shall be considered to have the same value as
ij1 ij2
the other repeat in accordance with the least squares method.
5.5.2 Both repeat values missing or rejected
5.5.2.1 General
If both the repeat values are missing, estimates of a (= y + y ) shall be made by forming the
ij ij1 ij2
laboratories × samples interaction sum of squares, including the missing values of the totals of the
laboratories/samples pairs of results as
...


NORME ISO
INTERNATIONALE 4259
Troisième édition
2006-08-01
Produits pétroliers — Détermination
et application des valeurs de fidélité
relatives aux méthodes d'essai
Petroleum products — Determination and application of precision data
in relation to methods of test

Numéro de référence
©
ISO 2006
PDF – Exonération de responsabilité
Le présent fichier PDF peut contenir des polices de caractères intégrées. Conformément aux conditions de licence d'Adobe, ce fichier
peut être imprimé ou visualisé, mais ne doit pas être modifié à moins que l'ordinateur employé à cet effet ne bénéficie d'une licence
autorisant l'utilisation de ces polices et que celles-ci y soient installées. Lors du téléchargement de ce fichier, les parties concernées
acceptent de fait la responsabilité de ne pas enfreindre les conditions de licence d'Adobe. Le Secrétariat central de l'ISO décline toute
responsabilité en la matière.
Adobe est une marque déposée d'Adobe Systems Incorporated.
Les détails relatifs aux produits logiciels utilisés pour la création du présent fichier PDF sont disponibles dans la rubrique General Info
du fichier; les paramètres de création PDF ont été optimisés pour l'impression. Toutes les mesures ont été prises pour garantir
l'exploitation de ce fichier par les comités membres de l'ISO. Dans le cas peu probable où surviendrait un problème d'utilisation,
veuillez en informer le Secrétariat central à l'adresse donnée ci-dessous.

©  ISO 2006
Droits de reproduction réservés. Sauf prescription différente, aucune partie de cette publication ne peut être reproduite ni utilisée sous
quelque forme que ce soit et par aucun procédé, électronique ou mécanique, y compris la photocopie et les microfilms, sans l'accord écrit
de l'ISO à l'adresse ci-après ou du comité membre de l'ISO dans le pays du demandeur.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax. + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Publié en Suisse
ii © ISO 2006 – Tous droits réservés

Sommaire Page
Avant-propos. v
Introduction . vi
1 Domaine d'application. 1
2 Références normatives . 1
3 Termes et définitions. 1
4 Étapes de l'organisation d'un programme d'essai interlaboratoires pour la détermination
de la fidélité d'une méthode d'essai . 4
4.1 Généralités . 4
4.2 Préparation d'un projet de méthode d'essai.5
4.3 Organisation d'un programme pilote avec au moins deux laboratoires. 5
4.4 Organisation d'un programme interlaboratoires. 5
4.5 Exécution du programme interlaboratoires. 6
5 Examen des résultats de l'essai interlaboratoires du point de vue de l'uniformité et des
valeurs aberrantes. 6
5.1 Généralités . 6
5.2 Transformation des données . 7
5.3 Recherche des valeurs aberrantes . 8
5.4 Rejet de tous les résultats concernant un échantillon . 11
5.5 Estimation des valeurs manquantes ou rejetées . 13
5.6 Test de rejet des résultats d'un laboratoires aberrant. 14
5.7 Confirmation de la transformation sélectionnée. 15
6 Analyse de la variance, calcul et expression des estimations de fidélité . 15
6.1 Généralités . 15
6.2 Analyse de la variance . 15
6.3 Espérance des carrés moyens et calcul des estimations de fidélité . 18
6.4 Expression de l’estimation de la fidélité dans une méthode d'essai . 21
7 Signification de la répétabilité (r) et de la reproductibilité (R). 22
7.1 Généralités . 22
7.2 Répétabilité, r . 22
7.3 Reproductibilité, R . 23
8 Spécifications. 25
8.1 But des spécifications. 25
8.2 Établissement des limites de spécifications en liaison avec la fidélité. 25
9 Contrôle de qualité et spécifications. 26
9.1 Généralités . 26
9.2 Marge de contrôle du fournisseur. 26
9.3 Marge de contrôle du réceptionnaire. 27
10 Règles de conformité et de non-conformité, en cas de désaccord. 27
Annexe A (normative) Détermination du nombre d'échantillons requis. 29
Annexe B (informative) Établissement de l'équation de calcul du nombre d'échantillons requis . 31
Annexe C (normative) Notation et essais . 33
Annexe D (normative) Exemples de résultats d'essais pour la détermination de l'indice de brome
et tableaux statistiques . 38
Annexe E (normative) Types de dépendance et transformations correspondantes . 47
Annexe F (normative) Analyse de régression linéaire pondérée . 51
Annexe G (normative) Règles d'arrondissage des résultats . 58
Annexe H (informative) Explication des équations données dans l'Article 7 . 59
Annexe I (informative) Spécifications relatives à un degré spécifié de caractère critique. 61
Bibliographie . 64

iv © ISO 2006 – Tous droits réservés

Avant-propos
L'ISO (Organisation internationale de normalisation) est une fédération mondiale d'organismes nationaux de
normalisation (comités membres de l'ISO). L'élaboration des Normes internationales est en général confiée
aux comités techniques de l'ISO. Chaque comité membre intéressé par une étude a le droit de faire partie du
comité technique créé à cet effet. Les organisations internationales, gouvernementales et non
gouvernementales, en liaison avec l'ISO participent également aux travaux. L'ISO collabore étroitement avec
la Commission électrotechnique internationale (CEI) en ce qui concerne la normalisation électrotechnique.
Les Normes internationales sont rédigées conformément aux règles données dans les Directives ISO/CEI,
Partie 2.
La tâche principale des comités techniques est d'élaborer les Normes internationales. Les projets de Normes
internationales adoptés par les comités techniques sont soumis aux comités membres pour vote. Leur
publication comme Normes internationales requiert l'approbation de 75 % au moins des comités membres
votants.
L'attention est appelée sur le fait que certains des éléments du présent document peuvent faire l'objet de
droits de propriété intellectuelle ou de droits analogues. L'ISO ne saurait être tenue pour responsable de ne
pas avoir identifié de tels droits de propriété et averti de leur existence.
L'ISO 4259 a été élaborée par le comité technique ISO/TC 28, Produits pétroliers et lubrifiants.
Cette troisième édition annule et remplace la deuxième édition (ISO 4259:1992), dont les Articles 1, 5, 7, C.7,
E.2 et F.3 et les paragraphes 4.2, 5.2, 6.3.2, 6.3.3.1, 6.3.3.3, 6.4, 8.2, 10.2, 10.4 et 10.5 ont fait l'objet d'une
révision technique. Elle incorpore également le Rectificatif technique ISO 4259:1992/Cor.1:1993.

Introduction
Pour les besoins de contrôle de qualité et pour vérifier leur conformité aux spécifications, les caractéristiques
des produits pétroliers commerciaux sont contrôlées au moyen de méthodes d'essai normalisées de
laboratoires. Deux ou plusieurs déterminations de la même caractéristique d'un échantillon donné, selon une
méthode d'essai donnée, ne donnent généralement pas exactement le même résultat. Il est donc nécessaire
de tenir compte correctement de ce fait, en parvenant à des estimations fondées sur les statistiques de la
fidélité d'une méthode, qui constituent une mesure objective du degré de concordance attendu entre deux ou
plusieurs résultats obtenus dans des conditions données.
[11]
L’ISO 4259 fait référence à l’ISO 3534-2 , qui donne une définition différente de la valeur vraie (voir 3.26).
L’ISO 4259 fait aussi référence à l’ISO 5725-2, qui est nécessaire pour l’estimation de la fidélité dans des
circonstances particulières et inhabituelles (voir 5.2).

vi © ISO 2006 – Tous droits réservés

NORME INTERNATIONALE ISO 4259:2006(F)

Produits pétroliers — Détermination et application des valeurs
de fidélité relatives aux méthodes d'essai
1 Domaine d'application
La présente Norme internationale traite du calcul des estimations de fidélité et de leur application aux
spécifications. En particulier, elle contient les définitions des termes statistiques concernés (Article 3), les
procédures à suivre dans l'organisation d'un programme d'essai interlaboratoires destiné à déterminer la
fidélité d'une méthode d'essai (Article 4), la méthode de calcul de la fidélité à partir des résultats d'un tel
programme (Article 5 et Article 6) et la procédure à suivre dans l'interprétation des résultats de laboratoires, à
la lumière de la fidélité des méthodes de test et des limites fixées dans les spécifications (Articles 7 à 10).
Les procédures de la présente Norme internationale ont été conçues spécifiquement pour les produits
pétroliers et leurs produits connexes, qui sont normalement homogènes. Cependant, on reconnaît que les
procédures décrites dans la présente Norme internationale peuvent aussi s’appliquer à d’autres types de
produits homogènes. Des contrôles attentifs s’avèrent nécessaires avant d’appliquer la présente Norme
internationale à des produits pour lesquels la présomption d’homogénéité peut être mise en question.
2 Références normatives
Les documents de référence suivants sont indispensables pour l'application du présent document. Pour les
références datées, seule l'édition citée s'applique. Pour les références non datées, la dernière édition du
document de référence s'applique (y compris les éventuels amendements).
ISO 5725-2:1994, Exactitude (justesse et fidélité) des résultats et méthodes de mesure — Partie 2: Méthode
de base pour la détermination de la répétabilité et de la reproductibilité d'une méthode de mesure normalisée
3 Termes et définitions
Pour les besoins du présent document, les termes et définitions suivants s'appliquent.
3.1
analyse de variance
technique qui permet de décomposer la variance totale d'une méthode en ses différents facteurs composants
3.2
variance interlaboratoires
élément de la variance totale dû aux différences entre les valeurs moyennes des différents laboratoires
NOTE 1 Lorsque des résultats obtenus par plus d'un laboratoires sont comparés, la dispersion est normalement plus
importante que lorsque le même nombre d'essais est effectué par un seul laboratoires, et il y a des écarts entre les
moyennes obtenues par les différents laboratoires. Ces écarts donnent lieu à la variance interlaboratoires.
NOTE 2 Il existe une définition correspondante pour la variance entre opérateurs.
NOTE 3 Le terme «interlaboratoires» est souvent abrégé en «laboratoires» lorsqu'il est utilisé pour qualifier des
paramètres représentatifs de la dispersion de la population de résultats, par exemple sous forme de «variance de
laboratoires».
3.3
biais
différence entre la valeur vraie (associée à la méthode d'essai) et la valeur connue, lorsque celle-ci est
disponible
NOTE Pour une définition de la valeur vraie et de la valeur connue, voir 3.26 et 3.8, respectivement.
3.4
codage aveugle
attribution d'un numéro différent pour chaque échantillon, afin qu'aucune autre identification ou information sur
les échantillons ne soit donnée à l'opérateur
3.5
échantillon de contrôle
échantillon prélevé au lieu où le produit est échangé, c'est-à-dire où la responsabilité de la qualité du produit
passe du fournisseur au réceptionnaire
3.6
degrés de liberté
diviseur utilisé dans le calcul de la variance; une unité de moins que le nombre de résultats indépendants
NOTE La définition n'est strictement applicable que dans les cas les plus simples. Des définitions plus complètes
sont en dehors de l'objet de la présente Norme internationale.
3.7
détermination
exécution de la série d'opérations spécifiées dans la méthode d'essai et permettant d'obtenir une seule valeur
3.8
valeur connue
valeur quantitative réelle découlant de la préparation de l'échantillon
NOTE La valeur connue n'existe pas toujours, par exemple dans le cas d'essais empiriques tels que la détermination
du point d'éclair.
3.9
moyenne
moyenne arithmétique
pour une série donnée de résultats, somme des résultats divisée par leur nombre
3.10
moyenne des carrés
somme des carrés divisée par le nombre de degrés de liberté
3.11
distribution normale
distribution de probabilité d'une variable aléatoire continue x, telle que si x est un nombre réel quelconque, la
densité de probabilité est:
⎡⎤
11⎛⎞x − µ
fx()=−exp⎢⎥,−∞ ⎜⎟
2 σ
σ 2π⎢⎥
⎝⎠
⎣⎦
NOTE µ est la valeur vraie et σ est l'écart-type de la distribution normale (σ > 0).
3.12
opérateur
personne qui effectue normalement et régulièrement un essai particulier
2 © ISO 2006 – Tous droits réservés

3.13
valeur aberrante
résultat dont la valeur est suffisamment éloignée des autres résultats pour qu'il ne soit pas considéré comme
faisant partie de l'ensemble des résultats
3.14
fidélité
étroitesse de l'accord entre les résultats obtenus en appliquant le procédé expérimental à plusieurs reprises
sur des produits identiques et dans des conditions déterminées
NOTE Le procédé est d'autant plus fidèle que la partie aléatoire des erreurs expérimentales qui affectent les résultats
est moindre.
3.15
erreur aléatoire
possibilité d'erreur dans tout essai en dépit du meilleur contrôle possible des variables
3.16
réceptionnaire
toute personne physique ou morale qui reçoit ou accepte le produit livré par le fournisseur
3.17
répétabilité
〈qualitativement〉 étroitesse de l'accord entre des résultats indépendants obtenus dans l'exécution normale et
correcte de la même méthode sur un produit identique soumis à l'essai, dans les mêmes conditions (même
opérateur, même appareillage, même laboratoires et courts intervalles de temps)
NOTE Les paramètres représentatifs de la dispersion de la population qui peut être associée au résultat sont
qualifiés du terme répétabilité, par exemple écart-type de répétabilité ou variance de répétabilité. Il convient de ne pas
confondre le terme «répétabilité» avec le terme «entre répétitions» ou «répétitions» lorsqu'il est utilisé de cette manière
(voir 3.19). La répétabilité se réfère à l'état de variabilité aléatoire minimale des résultats. Le temps pendant lequel des
résultats répétés doivent être obtenus sera donc suffisamment court pour exclure les erreurs qui dépendent du temps, par
exemple les erreurs dues à l'ambiance et à l'étalonnage.
3.18
répétabilité
〈quantitativement〉 valeur égale ou au dessous de laquelle la différence absolue entre deux résultats
individuels obtenus dans les conditions spécifiées est attendue, pour une probabilité de 95 %
NOTE Pour des détails sur les conditions spécifiées, voir 3.17.
3.19
répétition
exécution d'une méthode d'essai plus d'une fois pour améliorer la fidélité et obtenir une meilleure estimation
de l'erreur de détermination
NOTE Il convient de distinguer une répétition d'une redétermination dans la mesure où le premier terme implique que
les essais sont effectués à un seul endroit et, autant que possible, sur une seule période de temps. Les paramètres
représentatifs de la dispersion de la population qui peuvent être associés avec des expériences répétées sont qualifiés du
terme «entre répétitions» ou, sous une forme abrégée, de «répétitions», par exemple «écart-type des répétitions».
3.20
reproductibilité
〈qualitativement〉 étroitesse de l'accord entre des résultats individuels obtenus dans l'exécution normale et
correcte de la même méthode sur un produit identique soumis à l'essai, mais dans des conditions différentes
(opérateurs différents, appareillages différents et laboratoires différents)
NOTE Les paramètres représentatifs de la dispersion de la population qui peuvent être associés aux résultats sont
qualifiés du terme «reproductibilité», par exemple écart-type de reproductibilité ou variance de reproductibilité
3.21
reproductibilité
〈quantitativement〉 valeur à laquelle la différence absolue entre deux résultats individuels sur un produit
identique obtenus par des opérateurs différents dans des laboratoires différents, en utilisant la méthode
d'essai normalisée, doit rester égale ou inférieure, pour une probabilité spécifiée de 95 %
3.22
résultat
valeur finale obtenue en suivant le mode opératoire complet de la méthode d'essai, cette valeur pouvant être
obtenue au moyen d'une seule ou de plusieurs déterminations suivant les instructions de la méthode
NOTE Il est admis que le résultat est arrondi conformément à la procédure spécifiée dans l'Annexe G.
3.23
écart-type
mesure de la dispersion d'une série de résultats autour de leur moyenne, égale à la racine carrée positive de
la variance et estimée par la racine carrée positive de la moyenne des carrés
3.24
somme des carrés
somme des carrés de la différence entre une série de résultats et leur moyenne
3.25
fournisseur
toute personne physique ou morale responsable de la qualité d'un produit juste avant qu'il ne soit pris en
charge par le réceptionnaire
3.26
valeur vraie
pour les besoins pratiques, valeur vers laquelle tend la moyenne des résultats individuels obtenus par n
laboratoires, lorsque n tend vers l'infini
NOTE 1 Une valeur vraie ainsi définie est associée à chaque méthode d'essai particulière.
[11]
NOTE 2 Une définition différente et idéalisée est donnée dans l'ISO 3534-2 .
3.27
variance
moyenne des carrés de l'écart d'une variable aléatoire par rapport à sa moyenne, estimée par la moyenne
des carrés
4 Étapes de l'organisation d'un programme d'essai interlaboratoires pour la
détermination de la fidélité d'une méthode d'essai
4.1 Généralités
Les étapes de l'organisation d'un programme d'essai interlaboratoires sont les suivantes:
a) préparation d'un projet de méthode d'essai;
b) organisation d'un programme pilote avec au moins deux laboratoires;
c) organisation du programme interlaboratoires;
d) exécution du programme interlaboratoires.
Les quatre étapes sont décrites successivement de 4.2 à 4.5.
4 © ISO 2006 – Tous droits réservés

4.2 Préparation d'un projet de méthode d'essai
Celui-ci doit contenir tous les détails nécessaires pour l'exécution de l'essai et l'expression des résultats.
Toute condition susceptible d'avoir une influence sur les résultats doit être spécifiée.
À ce stade, seul le titre de l'article relatif à la fidélité est inclus dans le projet de méthode d’essai. Il est
recommandé que la limite inférieure du domaine d'application de la méthode d'essai ne soit pas inférieure à la
plus petite valeur testée dans le programme interlaboratoires et soit au moins plus grand de 2R au plus petit
résultat qui puisse être déterminé (voir 8.2), R étant l’estimation de la reproductibilité. De la même manière, il
est recommandé que la limite supérieure du domaine d'application de la méthode d'essai n'excède pas la
valeur maximale testée au cours du programme interlaboratoires et soit au moins supérieure de 2R au résultat
le plus élevé que l’on puisse atteindre.
4.3 Organisation d'un programme pilote avec au moins deux laboratoires
Un programme pilote est nécessaire pour les raisons suivantes:
a) pour vérifier les détails de l'exécution de l'essai;
b) pour déterminer dans quelle mesure les opérateurs peuvent observer correctement les instructions de la
méthode;
c) pour contrôler les instructions concernant les échantillons;
d) pour estimer grossièrement la fidélité de l'essai.
Au moins deux échantillons sont nécessaires, couvrant la plage de résultats pour laquelle la méthode d'essai
est conçue. Cependant, au moins douze combinaisons laboratoires/échantillon doivent être incluses. Chaque
échantillon est essayé deux fois par chaque laboratoires dans les conditions de répétabilité. Si des omissions
ou des imprécisions dans le projet de méthode d’essai sont révélées, elles doivent alors être corrigées. Les
résultats doivent être analysés sous l'angle du biais et de la fidélité: si l'un ou l'autre paraît trop important, des
modifications à la méthode d’essai doivent être envisagées.
4.4 Organisation d'un programme interlaboratoires
Celui-ci doit recueillir la participation d'au moins cinq laboratoires, mais il est préférable de dépasser ce
nombre pour réduire le nombre d'échantillons requis.
Le nombre d'échantillons doit être suffisant pour couvrir la plage de la propriété mesurée et rendre sûres les
estimations de fidélité. Si une quelconque variation de la fidélité avec le niveau des résultats est observée
dans le programme pilote, il devient nécessaire d'utiliser au moins cinq échantillons dans le programme
interlaboratoires. Dans tous les cas, il est nécessaire d'obtenir au moins 30 degrés de liberté pour la
répétabilité et la reproductibilité. Pour la répétabilité, cela signifie un total de 30 paires de résultats dans le
programme.
Pour la reproductibilité, le Tableau A.1 donne le nombre minimal d'échantillons requis en fonction de L, P et Q,
où L est le nombre de laboratoires participants et P et Q sont les rapports des estimations des composantes
de variance obtenues dans le programme pilote. Spécifiquement, P est le rapport de la composante
interaction à la composante répétitions et Q est le rapport de la composante laboratoires à la composante
répétitions. L'Annexe B donne le calcul de la formule utilisée. Si Q est beaucoup plus grand que P, les
30 degrés de liberté ne peuvent être atteints; les entrées vides dans le Tableau A.1 correspondent à cette
situation ou à une situation proche (c'est-à-dire lorsque plus de 20 échantillons sont nécessaires). Dans ces
cas, il y a vraisemblablement un biais significatif entre laboratoires.
4.5 Exécution du programme interlaboratoires
Une personne doit être responsable du programme entier, depuis la distribution des textes de la méthode
d’essai et des échantillons jusqu'à l'évaluation finale des résultats. Elle doit bien connaître la méthode d’essai,
mais ne doit pas prendre part personnellement aux essais.
Le texte de la méthode d’essai doit être diffusé à tous les laboratoires pour leur permettre de soulever
d'éventuelles questions avant le début des essais. Si un laboratoires désire pratiquer la méthode à l'avance,
cela doit être fait sur des échantillons autres que ceux utilisés dans le programme.
Les échantillons doivent être réunis, divisés et distribués par l'organisateur, qui doit également conserver une
réserve de chaque échantillon pour les cas urgents. Il est de la plus haute importance que les parties
fractionnées pour chaque laboratoires soient homogènes. Celles-ci doivent être codées avant la distribution,
et les informations suivantes doivent accompagner leur envoi.
a) La méthode (ou le projet de méthode) choisie pour les essai.
b) Les instructions pour la manipulation et le stockage des échantillons.
c) L'ordre dans lequel les échantillons doivent être soumis à l'essai (un ordre aléatoire différent pour chaque
laboratoires).
d) L'indication que deux résultats doivent être obtenus consécutivement sur chaque échantillon, par le
même opérateur avec le même appareillage. Pour des raisons statistiques, il est impératif que les deux
résultats soient obtenus indépendamment l'un de l'autre, c'est-à-dire que le second résultat ne soit pas
influencé par la connaissance du premier. Si cela est considéré comme impossible à atteindre par
l'opérateur concerné, les paires de résultats doivent être dans ce cas obtenues à l'aveugle, mais en
s'assurant que les essais sont exécutés sur une période de temps courte.
e) La période durant laquelle des résultats répétés doivent être obtenus et la période durant laquelle tous
les échantillons doivent être soumis à l'essai.
f) Un formulaire en blanc pour le report des résultats. Pour chaque échantillon, il doit être prévu la date de
l'essai, les deux résultats et toute circonstance inhabituelle. Le degré d'exactitude pour l'expression des
résultats doit être spécifié.
g) L'indication que l'essai doit être exécuté dans des conditions normales, par des opérateurs ayant une
bonne expérience mais pas de connaissances particulières; et que la durée de l'essai doit être conforme
à ce qui se fait normalement.
Les opérateurs du programme pilote peuvent prendre part au programme interlaboratoires. Si le surcroît
d'expérience qu'ils ont acquis lors de l'essai de quelques échantillons supplémentaires produit un effet notable,
cela devrait servir d'avertissement sur le fait que la méthode d’essais n'est pas satisfaisante. Ils doivent être
identifiés dans le compte rendu des résultats de sorte que tout effet puisse être noté.
5 Examen des résultats de l'essai interlaboratoires du point de vue de l'uniformité
et des valeurs aberrantes
5.1 Généralités
Les paragraphes 5.2 à 5.7 spécifient les procédures d'examen des résultats exprimés dans un essai circulaire
organisé dans un but statistique (voir l’Article 4) pour:
a) établir l'indépendance ou la dépendance de la fidélité et le niveau des résultats;
b) établir l'uniformité de la fidélité d'un laboratoires à un autre;
c) détecter la présence de valeurs aberrantes.
6 © ISO 2006 – Tous droits réservés

Les procédures sont décrites en termes mathématiques basés sur la notation de l'Annexe C et illustrées par
des références aux données de l'exemple (calcul de l'indice de brome) exposé dans l'Annexe D.
De 5.2 à 5.7 (et dans l'Article 6), les procédures à employer sont d'abord spécifiées, puis illustrées par un
exemple en utilisant les résultats donnés dans l'Annexe D.
Il est supposé, dans la suite du présent article, que tous les résultats soit appartiennent à une distribution
normale unique, soit sont susceptibles d'être transformés en une telle distribution (voir 5.2). Les autres cas
(qui sont rares) nécessitent un traitement différent qui n'est pas dans l'objet de la présente Norme
internationale. Se reporter à la Référence [8] en Bibliographie pour l'essai statistique de normalité.
Bien que les procédures illustrées ici soient sous une forme permettant le calcul manuel, il est fortement
recommandé d'utiliser un calculateur électronique pour stocker et analyser les résultats des essais
interlaboratoires sur la base des procédures de la présente Norme internationale (voir [9] par exemple).
5.2 Transformation des données
5.2.1 Généralités
Dans de nombreuses méthodes d'essai, la fidélité dépend du niveau du résultat de l'essai, en sorte que la
dispersion des résultats diffère d'un échantillon à un autre. La méthode d'analyse décrite dans la présente
Norme internationale suppose qu'il n'en est pas ainsi, et la situation est corrigée, si nécessaire, par une
transformation.
Les écarts-types D des laboratoires et les écarts-types d des répétitions pour l’échantillon j (voir l’Annexe C)
j j
sont calculés et reportés dans des graphiques différents par rapport aux moyennes m des échantillons. Si les
j
courbes ainsi obtenues correspondent approximativement à deux droites parallèles à l'axe m, il n'est pas
nécessaire de procéder à une transformation. Mais si les points sont disposés approximativement suivant des
courbes de la forme D = f (m) et d = f (m), une transformation est nécessaire.
1 2
Les relations D = f (m) et d = f (m) ne sont généralement pas identiques. Les procédures statistiques de la
1 2
présente Norme internationale exigent toutefois que la même transformation soit applicable en même temps à
la répétabilité et à la reproductibilité. Pour cette raison, les deux relations sont combinées dans une relation
de dépendance unique, D = f(m) (où D inclut à présent d), en introduisant une variable fictive T. Cela tient
compte de la différence entre les relations, s'il y en a, et fournit un moyen de tester cette différence (voir F.1).
La relation simple D = f(m) est évaluée au mieux par une analyse de régression linéaire pondérée, même si
dans la plupart des cas une régression non pondérée constitue une approximation suffisante. La dérivation
des pondérations est décrite dans l'Article F.2 et la procédure de calcul pour l'analyse de régression est
décrite dans l'Article F.3. Des formes types de dépendance D = f(m) sont données dans l'Article E.1. Elles
sont toutes exprimées en fonction de paramètres de transformation B et B .
L'estimation de B et B et la procédure de transformation qui suit sont résumées dans l'Article E.2. Celui-ci
comporte les essais statistiques pour la signification de la régression (c'est-à-dire, savoir si la relation D = f(m)
est parallèle à l'axe m) et pour la différence entre les relations de répétabilité et de reproductibilité basées sur
une probabilité à 95 %. Si une telle différence est constatée, ou si aucune transformation appropriée n'existe,
les procédures alternatives échantillon par échantillon de l'ISO 5725-2 doivent être utilisées. En pareil cas, il
n’est pas possible de tester les biais de laboratoires de tous les échantillons (voir 5.6) ni d'évaluer la
composante d'interaction de variance (voir 6.2).
S'il apparaît, avec une probabilité de 95 %, qu'il existe une régression significative de la forme D = f(m), la
transformation adéquate y = F(x), où x est le résultat noté, est donnée par la formule:
dx
Fx =K (2)
()

f x
()
où K est une constante. Dans ce cas, tous les résultats doivent être transformés en conséquence et le reste
de l'analyse exécuté en fonction des résultats transformés. Des transformations types sont données en E.1.
Le choix d'une transformation est difficile et ne peut faire l'objet de règles formelles. Une assistance statistique
qualifiée peut être nécessaire dans certains cas particuliers. La présence de valeurs aberrantes peut
influencer le jugement quant au type de transformation nécessaire éventuellement (voir 5.7).
5.2.2 Exemple
Le Tableau 1 reprend les valeurs de m, D et d pour les huit échantillons de l'exemple donné dans l'Annexe D,
exprimé avec trois chiffres significatifs. Les degrés de liberté correspondants sont indiqués entre parenthèses.
Tableau 1
Numéro
3 8 1 4 5 6 2 7
d'échantillon
m 0,756 1,22 2,15 3,64 10,9 48,2 65,4 114
D 0,066 9 (14) 0,159 (9) 0,729 (8) 0,211 (11) 0,291 (9) 1,50 (9) 2,22 (9) 2,93 (9)
d
0,050 0 (9) 0,057 2 (9) 0,127 (9) 0,116 (9) 0,094 3 (9) 0,527 (9) 0,818 (9) 0,935 (9)
De l'examen des nombres du Tableau 1, il apparaît qu'à la fois D et d augmentent avec m, mais d'autant
moins rapidement que m est plus grand. Si l'on reporte ces nombres sur un papier log-log (c'est-à-dire si l'on
réalise un graphique de log D et log d en fonction de log m), on constate que les points sont disposés
approximativement suivant deux lignes droites (voir Figure F.1). Les calculs donnés comme exemple dans
l'Article F.4 montrent que les pentes de ces droites sont les mêmes, leur valeur estimée étant de 0,638. Si l'on
tient compte des erreurs d'estimation de ces pentes, on peut admettre qu'il s'agit de droites de pente 2/3.
Par conséquent, la même transformation convient pour la répétabilité et la reproductibilité, et est donnée par
la formule:
−21
xxd3=x (3)

Puisqu'on peut négliger le coefficient constant, la transformation se réduit à l'extraction des racines cubiques
des résultats obtenus (indices de brome). Cette opération fournit les données transformées figurant dans le
Tableau D.2, dans lequel les racines cubiques sont données avec trois décimales.
5.3 Recherche des valeurs aberrantes
5.3.1 Généralités
Les résultats notés ou, si l'on a décidé qu'une transformation était nécessaire, les résultats transformés,
doivent être étudiés pour rechercher les valeurs aberrantes. Ce sont des valeurs si différentes des autres qu'il
peut être admis qu'elles proviennent d'une quelconque erreur dans l'application de la méthode d’essai ou de
l'essai d'un mauvais échantillon. On peut utiliser de nombreux tests d'élimination et faire varier les seuils de
signification associés, mais on a constaté que ceux donnés ci-après conviennent à l'objet de la présente
Norme internationale. Cette recherche des valeurs aberrantes suppose, dans tous les cas, une distribution
normale des erreurs (voir 5.1).
5.3.2 Uniformité de la répétabilité
5.3.2.1 Généralités
Le premier test de recherche des valeurs aberrantes a pour but de détecter un résultat discordant dans
[1]
une paire de résultats répétés. Ce test implique le calcul de e sur toutes les combinaisons
ij
laboratoire/échantillon. On applique ensuite le critère de Cochran au niveau de signification 1 % au rapport de
la plus grande de ces e valeurs à leur somme (voir l’Article C.5). Si sa valeur dépasse la valeur donnée
ij
8 © ISO 2006 – Tous droits réservés

dans le Tableau D.3 correspondant à un degré de liberté, n étant le nombre de paires soumises à
comparaison, il faut rejeter l'élément de la paire qui est le plus éloigné de la moyenne de l'échantillon et
répéter le processus en réduisant n de 1 unité, jusqu'à ce qu'il n'y ait plus de rejet. Dans certains cas, ce test
fait «boule de neige» et conduit à une proportion de rejets inacceptable (par exemple plus de 10 %). Si c'est
le cas, il faut abandonner ce test de rejet et conserver quelques-uns ou la totalité des résultats rejetés. Une
décision arbitraire basée sur un jugement est alors nécessaire.
5.3.2.2 Exemple
Dans le cas de l'exemple donné dans l'Annexe D, les différences absolues entre les résultats répétés
transformés, c'est-à-dire entre les paires de nombres du Tableau D.2, en unités de la troisième décimale,
figurent dans le Tableau 2.
Tableau 2
Échantillon
Laboratoires
1 2 3 4 5 6 7 8
A
42 21 7 13 7 10 8 0
B
23 12 12 0 7 9 3 0
C
0 6 0 0 7 8 4 0
D 14 6 0 13 0 8 9 32
E 65 4 0 0 14 5 7 28
F 23 20 34 29 20 30 43 0
G 62 4 78 0 0 16 18 56
H 44 20 29 44 0 27 4 32
J 0 59 0 40 0 30 26 0
L'écart le plus grand est de 0,078 pour le laboratoires G sur l'échantillon 3. La somme des carrés de tous les
écarts est:
22 22
0,042++0,021 .+ 0,026+ 0= 0,043 9
Par suite, le rapport à comparer au critère de Cochran est:
0,078
= 0,138
0,043 9
Il y a 72 paires et, comme cela ressort du Tableau D.3, le critère pour 80 paires est 0,170 9, ce rapport n'est
pas significatif.
5.3.3 Uniformité de la reproductibilité
5.3.3.1 Généralités
Les tests suivants de valeurs aberrantes s'attachent à établir l'uniformité de la reproductibilité estimée. Elles
ont pour but de déceler soit une paire de résultats discordants obtenus par un laboratoires sur un échantillon
particulier, soit un groupe de résultats discordants provenant d'un laboratoires sur tous les échantillons. Dans
[2]
ces deux cas, on peut utiliser le test de Hawkins .
Il s'agit de former pour chaque échantillon, puis pour les moyennes globales des laboratoires (voir 5.6), le
rapport du plus grand écart absolu de la moyenne des laboratoires par rapport à la moyenne des échantillons
(ou la moyenne globale) à la racine carrée de certaines sommes de carrés (voir C.6).
Le rapport correspondant à l'écart absolu le plus important doit être comparé aux valeurs critiques 1 %
données dans le Tableau D.4, où n est le nombre de cellules laboratoires/échantillon dans l'échantillon
concerné (ou le nombre de moyennes de laboratoires globales) et ν représente les degrés de liberté pour la
somme des carrés qui s'ajoute à celle correspondant à l'échantillon en question. Dans l'essai pour les cellules
laboratoires/échantillon, ν se réfère à d'autres échantillons, mais est égal à zéro dans l'essai pour les
moyennes de laboratoires globales.
Si une valeur significative se manifeste pour un ou plusieurs échantillons, les valeurs extrêmes
correspondantes doivent être omises et le processus répété. Si des valeurs extrêmes quelconques sont
trouvées parmi les sommes des résultats de chaque laboratoires correspondant, il faut rejeter tous les
résultats provenant du ou des laboratoires.
Si l'essai fait «boule de neige», conduisant ainsi à une proportion de rejets inacceptable (par exemple plus de
10 %), ce test de rejet doit être abandonné et quelques-uns ou la totalité des résultats qui avaient été rejetés
doivent être retenus. Une décision arbitraire basée sur un jugement est nécessaire dans ce cas.
5.3.3.2 Exemple
L'application du test de Hawkins aux moyennes de cellule dans les échantillons est illustrée ci-après.
La première étape consiste à calculer les écarts des moyennes de cellule par rapport aux moyennes
d'échantillon respectives sur l'ensemble. Ils sont illustrés dans le Tableau 3, en unités de la troisième
décimale.
La somme des carrés des écarts est ensuite calculée pour chaque échantillon. Ces sommes sont illustrées
dans le Tableau 3 en unités de la troisième décimale.
Tableau 3
Échantillon
Laboratoires
1 2 3 4 5 6 7 8
A 20 8 14 15 10 48 6 3
B 75 7 20 9 10 47 6 3
C 64 35 3 20 30 4 22 25
D 314 33 18 42 7 39 80 50
E 32 32 30 9 7 18 18 39
F 75 97 31 20 30 8 74 53
G 10 34 32 20 20 61 9 62
H 42 13 4 42 13 21 8 50
J 1 28 22 29 14 8 10 53
Somme des carrés 117 15 4 6 3 11 13 17

10 © ISO 2006 – Tous droits réservés

La cellule à tester est celle qui présente la déviation la plus extrême. Cela a été obtenu par le laboratoires D à
partir de l'échantillon 1. Le rapport approprié de Hawkins pour l'essai est donc:
0,314
B*== 0,728 1
0,117++0,015 .+ 0,017
La valeur critique, correspondant à n = 9 cellules dans l'échantillon 1 et ν = 56 degrés de liberté pour les
autres échantillons, est interpolée à partir du Tableau D.4 comme étant 0,372 9. La valeur du test est
supérieure à la valeur critique, en conséquence les résultats du laboratoires D pour l'échantillon 1 sont rejetés.
Comme il y a eu rejet, la valeur moyenne, les écarts et la somme des carrés sont recalculés pour
l'échantillon 1 et la procédure répétée. La cellule suivante à tester est celle obtenue par le laboratoires F à
partir de l'échantillon 2. Le rapport de Hawkins pour l'essai pour cette cellule est le suivant:
0,097
B*== 0,354 2
0,006++0,015 .+ 0,017
La valeur critique, correspondant à n = 9 cellules dans l'échantillon 2 et ν = 55 degrés de liberté, est interpolé
à partir du Tableau D.4 comme étant 0,375 6. Comme le rapport pour l'essai est inférieur à la valeur type, il
n'y a pas d'autres rejets.
5.4 Rejet de tous les résultats concernant un échantillon
5.4.1 Généralités
Il faut examiner l'écart-type des laboratoires et l'écart-type des répétitions en vue de déceler la présence
d'échantillons aberrants. Si l'on a procédé à une transformation ou si un rejet a été fait, il faut calculer de
nouveaux écarts-types.
Si l'écart-type pour l’un des échantillons est beaucoup trop grand, il faut l'examiner avec l'idée de rejeter les
résultats de cet échantillon.
Le critère de Cochran au niveau 1 % peut être utilisé lorsque les écarts-types sont basés sur le même nombre
de degrés de liberté. Cela implique le calcul du rapport de la plus grande somme de carrés correspondante
(laboratoires ou répétitions, selon le cas) vis-à-vis de leur total (voir l’Article C.5). Si ce rapport dépasse la
valeur critique donnée dans le Tableau D.3, n étant le nombre d'échantillons et ν les degrés de liberté, tous
les résultats de l'échantillon en question doivent être rejetés. En pareil cas, il convient de veiller à ce que
l'écart-type extrême ne soit pas dû à l'application d'une transformation inappropriée (voir 5.2) ou à des valeurs
aberrantes non détectées.
Il n'existe pas d'essai optimal lorsque les écarts-types sont basés sur différents degrés de liberté. Toutefois, le
rapport de l'écart le plus grand à celui obtenu à partir des échantillons restants suit une loi de F avec des
degrés de liberté ν et ν (voir l’Article C.7). Dans ce cas, ν représente les degrés
...


МЕЖДУНАРОДНЫЙ ISO
СТАНДАРТ
Третье издание
2006-08-01
Нефтепродукты. Определение и
применение данных прецизионности в
отношении методов испытания
Petroleum products — Determination and application of precision data
in relation to methods of test

Ответственность за подготовку русской версии несёт GOST R
(Российская Федерация) в соответствии со статьёй 18.1 Устава ISO
Ссылочный номер
©
ISO 2006
Отказ от ответственности при работе в PDF
Настоящий файл PDF может содержать интегрированные шрифты. В соответствии с условиями лицензирования, принятыми
фирмой Adobe, этот файл можно распечатать или смотреть на экране, но его нельзя изменить, пока не будет получена
лицензия на интегрированные шрифты и они не будут установлены на компьютере, на котором ведется редактирование. В
случае загрузки настоящего файла заинтересованные стороны принимают на себя ответственность за соблюдение
лицензионных условий фирмы Adobe. Центральный секретариат ISO не несет никакой ответственности в этом отношении.
Adobe - торговый знак фирмы Adobe Systems Incorporated.
Подробности, относящиеся к программным продуктам, использованные для создания настоящего файла PDF, можно найти в
рубрике General Info файла; параметры создания PDF были оптимизированы для печати. Были приняты во внимание все
меры предосторожности с тем, чтобы обеспечить пригодность настоящего файла для использования комитетами-членами
ISO. В редких случаях возникновения проблемы, связанной со сказанным выше, просьба проинформировать Центральный
секретариат по адресу, приведенному ниже.

ДОКУМЕНТ ЗАЩИЩЕН АВТОРСКИМ ПРАВОМ

©  ISO 2006
Все права сохраняются. Если не указано иное, никакую часть настоящей публикации нельзя копировать или использовать в
какой-либо форме или каким-либо электронным или механическим способом, включая фотокопии и микрофильмы, без
предварительного письменного согласия ISO, которое должно быть получено после запроса о разрешении, направленного по
адресу, приведенному ниже, или в комитет-член ISO в стране запрашивающей стороны.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Опубликовано в Швейцарии
ii © ISO 2006 – Все права сохраняются

Содержание Страница
Предисловие .v
Введение .vi
1 Область применения .1
2 Нормативные ссылки .1
3 Термины и определения .1
4 Этапы планирования межлабораторной программы испытаний для определения
прецизионности метода испытания .5
4.1 Общие положения .5
4.2 Подготовка проекта метода испытания.5
4.3 Планирование пилотной (пробной) программы с участием не менее двух
лабораторий .6
4.4 Планирование межлабораторной программы .6
4.5 Выполнение межлабораторной программы.6
5 Исследование результатов межлабораторных испытаний с целью проверки
однородности и выявления выбросов.7
5.1 Общие положения .7
5.2 Преобразование данных.8
5.3 Выявление аномальных результатов (выбросов).9
5.4 Браковка полного набора данных по пробе .12
5.5 Оценивание результатов, заменяющих потерянные и забракованные данные.14
5.6 Выявление выпадающих лабораторий .15
5.7 Подтверждение правильности выбранного преобразования .16
6 Дисперсионный анализ и вычисление оценок показателей прецизионности .16
6.1 Общие положения .16
6.2 Дисперсионный анализ.16
6.3 Математические ожидания средних квадратов и расчет оценок показателей
прецизионности .19
6.4 Выражение оценок показателей прецизионности метода испытания .22
7 Значение повторяемости (r) и воспроизводимости (R).23
7.1 Общие положения .23
7.2 Повторяемость (сходимость), r .23
7.3 Воспроизводимость, R.24
8 Спецификации (Технические условия на продукцию).26
8.1 Назначение спецификаций .26
8.2 Формирование граничных значений состава и свойства нефтепродуктов с учетом
прецизионности .26
9 Контроль качества продукции на соответствие требованиям спецификаций.27
9.1 Общие положения .27
9.2 Граничные значения результатов у поставщика.27
9.3 Граничные значения результатов у получателя .28
10 Процедура урегулирования разногласий.28
Приложение A (нормативное) Определение требуемого числа проб .30
Приложение B (информативное) Вывод уравнения для расчета требуемого числа проб .32
Приложение C (нормативное) Виды испытаний и обозначения .33
Приложение D (нормативное) Примеры результатов испытаний по определению бромного
числа и статистические таблицы . 38
Приложение E (нормативное) Виды зависимостей и соответствующие преобразования . 47
Приложение F (нормативное) Взвешенный линейный регрессионный анализ . 51
Приложение G (нормативное) Правила округления результатов . 58
Приложение H (информативное) Пояснение к уравнениям, приведенным в Разделе 7. 59
Приложение I (информативное) Уровень требований к качеству продукции в
спецификациях. 61
Библиография . 64

iv © ISO 2006 – Все права сохраняются

Предисловие
Международная организация по стандартизации (ISO) является всемирной федерацией национальных
организаций по стандартизации (комитетов-членов ISO). Разработка международных стандартов
обычно осуществляется техническими комитетами ISO. Каждый комитет-член, заинтересованный в
деятельности, для которой был создан технический комитет, имеет право быть представленным в этом
комитете. Международные правительственные и неправительственные организации, имеющие связи с
ISO, также принимают участие в работах. Что касается стандартизации в области электротехники, то
ISO работает в тесном сотрудничестве с Международной электротехнической комиссией (IEC).
Проекты международных стандартов разрабатываются в соответствии с правилами Директив ISO/IEC,
Часть 2.
Основная задача технических комитетов заключается в подготовке международных стандартов.
Проекты международных стандартов, принятые техническими комитетами, рассылаются комитетам-
членам на голосование. Их опубликование в качестве международных стандартов требует одобрения
не менее 75 % комитетов-членов, принимающих участие в голосовании.
Следует иметь в виду, что некоторые элементы настоящего международного стандарта могут быть
объектом патентных прав. ISO не может нести ответственность за идентификацию какого-либо одного
или всех патентных прав.
Международный стандарт ISO 4259 подготовлен Техническим комитетом ISO/TC 28, Нефтепродукты
и смазочные материалы.
Настоящее третье издание отменяет и заменяет второе издание (ISO 4259:1992), Разделы 1, 5, 7, С.7, Е.2 и F.3
и подразделы 4.2, 5.2, 6.3.2, 6.3.3.1, 6.3.3.3, 6.4, 8.2, 10.2, 10.4 и 10.5, подверглись техническому пересмотру.
Этот стандарт также включает Техническую поправку ISO 4259:1992/Cor.1:1993.
Введение
В целях контроля качества и проверки соответствия техническим требованиям спецификаций
показатели товарные нефтепродукты оценивают с помощью стандартизованных лабораторных
методов испытаний. Два или более измерений одного и того же свойства конкретной пробы любым
заданным методом испытаний (измерений) обычно не дают точно один и тот же результат. Поэтому
необходимо должным образом учитывать этот факт, принять статистически обоснованные оценки
показателей прецизионности метода, т.е. объективные меры согласованности, ожидаемой для двух
или более результатов, полученных в установленных условиях.
[11]
ISO 4259 дает ссылку на ISO 3534-1 , в котором приводится другое определение истинного значения
(см. 3.26). ISO 4259 также ссылается на ISO 5725-2, который требуется при особых или необычных
обстоятельствах (см. 5.2) с целью оценки прецизионности.

vi © ISO 2006 – Все права сохраняются

МЕЖДУНАРОДНЫЙ СТАНДАРТ ISO 4259:2006(R)

Нефтепродукты. Определение и применение данных
прецизионности в отношении методов испытания
1 Область применения
Данный международный стандарт охватывает расчет оценок прецизионности и их применение к
спецификациям (техническим условиям). В частности, стандарт содержит определения
соответствующих статистических терминов (Раздел 3), процедуры, которые необходимо выполнять
при планировании межлабораторной программы испытаний для определения прецизионности метода
испытания (Раздел 4), метод расчета прецизионности по результатам такой программы (Разделы 5 и
6), и процедуру, которой необходимо следовать при интерпретации лабораторных результатов как в
отношении прецизионности методов испытаний, так и пределов (граничных значений), установленных
в спецификациях (Разделы 7 – 10).
Процедуры, описанные в данном международном стандарте, предназначены конкретно для нефти и
нефтепродуктов, которые обычно являются гомогенными. Однако, процедуры, описанные в этом
стандарте, можно также применить к другим типам гомогенных продуктов. Тщательные исследования
необходимо провести, прежде чем применять настоящий международный стандарт к продуктам, для
которых предположение о гомогенности находится под вопросом.
2 Нормативные ссылки
Следующие нормативные документы являются обязательными для применения с настоящим
международным стандартом. Для жестких ссылок применяются только указанное по тексту издание.
Для плавающих ссылок необходимо использовать самое последнее издание нормативного ссылочного
документа (включая любые изменения).
ISO 5725-2:1994, Точность (правильность и прецизионность) методов и результатов измерения.
Часть 2. Основной метод определения повторяемости (сходимости) и воспроизводимости
стандартного метода измерения
3 Термины и определения
Применительно к настоящему стандарту используются следующие термины и определения.
3.1
дисперсионный анализ
analysis of variance
метод анализа, позволяющий разложить общую дисперсию на составляющие, обусловленные
действием определенных факторов
3.2
межлабораторная дисперсия
between-laboratory variance
элемент полной дисперсии, свойственный расхождению между средними значениями для разных
лабораторий
ПРИМЕЧАНИЕ 1 Если сопоставляют результаты, полученные более чем в одной лаборатории, разброс обычно
шире, чем если то же самое число испытаний выполняют в одной лаборатории, и существует некоторая вариация
между средними значениями, полученными разными лабораториями. Это вызывает увеличение межлабораторной
дисперсии, которая является составляющей общей дисперсии, обусловленное расхождением средних значений,
полученных в разных лабораториях.
ПРИМЕЧАНИЕ 2 Существует соответствующее определение дисперсии между операторами.
ПРИМЕЧАНИЕ 3 Термин “межлабораторный” часто сокращают до термина “лабораторный”, если его используют
для оценки репрезентативных представительных параметров дисперсии совокупности результатов, например,
“лабораторная дисперсия”.
3.3
смещение (отклонение)
bias
разность между истинным значением (относящимся к данному методу испытания) и известным
значением, если его можно определить
ПРИМЕЧАНИЕ Для определения “истинного значения” и “известного значения” см. 3.26 и 3.8, соответственно.
3.4
слепое кодирование
blind coding
присвоение разных номеров каждой из проб, при этом оператор не должен получать никакой другой
идентификации или информации о пробе
3.5
контрольная проба
check sample
проба, отобранная в том месте, где происходит передача продукта, т.е. там, где ответственность за
качество продукта переходит от поставщика к получателю
3.6
число степеней свободы
degrees of freedom
делитель, используемый при расчете дисперсии; на единицу меньше числа независимых результатов
ПРИМЕЧАНИЕ Это определение применяется только в простейших случаях. Исчерпывающие определения
выходят за рамки настоящего международного стандарта.
3.7
определение
determination
процесс выполнения серии операций, установленных в документе на метод испытаний, в ходе которых
получают единичное значение
3.8
известное значение
known value
фактическое количественное значение величины, установленное процедурой приготовления
ПРИМЕЧАНИЕ Известное значение не всегда существует (оно не может быть определено), например, для
эмпирических условных величин, таких как температура вспышки.
3.9
среднее
среднее арифметическое
mean
arithmetic mean
сумма результатов для данного ряда, деленная на их число
2 © ISO 2006 – Все права сохраняются

3.10
среднее квадратичное (средний квадрат)
mean square
сумма квадратов, деленная на число степеней свободы
3.11
нормальное распределение
normal distribution
распределение вероятностей непрерывной случайной переменной, х, такой, что если х любое
действительное число, плотность вероятности равна

11x − µ
fx=−exp,−∞ ()

2 σ
σ 2р


ПРИМЕЧАНИЕ µ является истинным значением, а σ среднеквадратическим отклонением данного нормального
распределения (σ > 0).
3.12
оператор
operator
сотрудник, который обычно и регулярно выполняет конкретное испытание
3.13
выброс
outlier
результат, достаточно далекий по величине от других результатов, и его нельзя рассматривать как
часть одного ряда
3.14
прецизионность
precision
степень близости друг к другу результатов, полученных при неоднократном выполнении процедуры
испытаний на идентичных материалах и в предписанных условиях
ПРИМЕЧАНИЕ Чем меньше случайная составляющая экспериментальной ошибки, тем выше точность метода.
3.15
случайная ошибка
random error
случайное отклонение, встречающееся во всех испытаниях, несмотря на самый строгий контроль
переменных
3.16
получатель
recipient
любое лицо или организация, которая получает или принимает продукцию от поставщика
3.17
повторяемость (сходимость)
repeatability
〈качественно〉 степень близости друг к другу независимых результатов, полученных при нормальном и
правильном выполнении одного и того же метода на идентичном испытуемом материале за короткий
промежуток времени в одних и тех же условиях испытания (один и тот же оператор, одно и то же
оборудование, одна и та же лаборатория)
ПРИМЕЧАНИЕ Представительные характеристики рассеяния генеральной совокупности, которые можно
связать с результатами испытаний, охватываются термином «повторяемость», например, стандартное
(среднеквадратическое) отклонение в условиях повторяемости или дисперсия в условиях повторяемости. Важно,
чтобы термин “повторяемость” не путали с терминами “между повторными” или “повторные испытания” (см. 3.19).
Повторяемость относят к состоянию, когда случайная изменчивость результатов минимальна. Период времени, в
течение которого повторяемые результаты должны быть получены, должен быть достаточно коротким, чтобы
исключить зависимые от времени ошибки, ошибки калибровки и ошибки, связанные с влиянием окружающей
среды.
3.18
повторяемость (сходимость)
repeatability
〈количественно〉 абсолютное значение, которое меньше или равно разности двух отдельных
результатов испытаний, полученных в заданных условиях, соответствующее доверительной
вероятности 95 %
ПРИМЕЧАНИЕ В отношении подробного описания заданных условий см. 3.17.
3.19
дублирование
replication
выполнение метода испытания более одного раза с целью улучшения прецизионности и получения
меньшей ошибки испытаний
ПРИМЕЧАНИЕ Дублирование следует отличать от обычного повторения, так как оно предполагает выполнение
повторных испытаний в одном месте и, по мере возможности, в течение одного периода времени.
представительные характеристики дисперсии совокупности, которую можно ассоциировать с повторными
экспериментами, обозначают термином “между повторами” или сокращенно “повторы”, например “стандартное
отклонение повторов”.
3.20
воспроизводимость
reproducibility
〈качественно〉 степень близости друг к другу отдельных результатов, полученных при нормальном и
правильном выполнении одного и того же метода на идентичном испытуемом материале, но в разных
условиях испытания (разные операторы, разное оборудование, разные лаборатории).
ПРИМЕЧАНИЕ Представительные характеристики дисперсии генеральной совокупности, которую можно
ассоциировать с результатами испытаний, охватываются термином “воспроизводимость”, например, стандартное
(среднеквадратическое) отклонение в условиях воспроизводимости или дисперсия в условиях воспроизводимости.
3.21
вопроизводимость
repproducibility
〈количественно〉 значение, которое меньше или равно абсолютной разности двух независимых результатов
испытаний, полученных на идентичном материале операторами в разных лабораториях с использованием
стандартизованного метода испытаний, соответствующее доверительной вероятности 95 %
3.22
результат
result
итоговое значение, полученное путем выполнения процедуры метода испытания; оно может быть
получено из одного определения или нескольких определений в зависимости от указаний метода
ПРИМЕЧАНИЕ Предполагается, что результат округляют в соответствии с процедурой, описанной в
Приложении G.
3.23
стандартное (среднеквадратическое) отклонение
standard deviation
мера рассеяния серии результатов относительно среднего значения, равная положительному
значению корня квадратного от среднего квадратичного (среднего квадрата)
4 © ISO 2006 – Все права сохраняются

3.24
сумма квадратов
sum of squares
сумма квадратов разностей серий результатов и их среднего значения
3.25
поставщик
supplier
лицо или организация, ответственная за качество продукции, прежде чем ее заберет получатель
3.26
истинное значение
true value
для практических целей, значение, к которому стремится среднее значение отдельных результатов,
полученных n лабораториями, при n, стремящемся к бесконечности
ПРИМЕЧАНИЕ 1 Такое истинное значение связано с конкретным методом испытания.
[11]
ПРИМЕЧАНИЕ 2 Другое идеализированное определение приводится в стандарте ISO 3534-2 .
3.27
дисперсия
variance
среднее значение квадратов отклонения случайной переменной от ее среднего, рассчитанного по
среднему квадрату
4 Этапы планирования межлабораторной программы испытаний для
определения прецизионности метода испытания
4.1 Общие положения
Этапы планирования межлабораторной программы испытаний следующие:
а) подготовка проекта метода испытания;
b) планирование экспериментальной (пробной) программы для не менее двух лабораторий,
с) планирование межлабораторной программы;
d) выполнение межлабораторной программы.
Эти четыре этапа описаны по очереди с 4.2 по 4.5.
4.2 Подготовка проекта метода испытания
Этот этап должен включать описание всех необходимых деталей для выполнений испытаний и форму
отчета о его результатах. Необходимо регламентировать любое условие, которое могло бы изменить
результаты испытания.
Раздел по прецизионности включает в проект метод испытания на данном этапе только как заголовок.
Рекомендуется, чтобы нижняя граница области применения данного метода испытания была не
меньше чем область самого низкого значения, измеренного в межлабораторной программе, и, по
крайней мере, на 2R больше, чем самый низкий достижимый результат (см. 8.2), где R – оценка
воспроизводимости. Аналогично, рекомендуется, чтобы верхняя граница области применения данного
метода испытания была не больше чем область самого высокого значения, измеренного в
межлабораторной программы, и, по крайней мере, на 2R меньше, чем самый высокий достижимый
результат.
4.3 Планирование пилотной (пробной) программы с участием не менее двух
лабораторий
Пилотная программа необходима по следующим причинам:
а) для проверки деталей при выполнении испытания;
b) чтобы выяснить, насколько грамотно операторы могут следовать инструкциям по методу
испытания;
с) чтобы проверить меры предосторожности при обращении с пробами;
d) чтобы провести предварительную оценку показателей прецизионности данного метода
испытания.
Требуется не менее двух проб, охватывающих диапазон определяемых значений, для которых
предназначен данный метод, однако, должно быть включено не менее 12 комбинаций
лаборатория/проба. Каждую пробу испытывают дважды в каждой лаборатории в условиях
повторяемости. Если в проекте метода испытания обнаруживаются какие-либо пропуски или
неточности, их необходимо исправить. Результаты необходимо проанализировать в отношении
отклонения (смещения) и прецизионности; если выявлены слишком большие нарушения, то
необходимо рассмотреть изменения к методу испытания.
4.4 Планирование межлабораторной программы
В испытаниях должно участвовать не менее пяти лабораторий, но предпочтительно, чтобы участниц
было больше, чтобы сократить число требуемых проб.
Число проб должно быть достаточным, чтобы охватить диапазон уровней измеряемого свойства
приблизительно через равноотстоящие интервалы и придать надежность оценкам прецизионности.
Если обнаружено, что прецизионность изменяется в зависимости от уровня значений результатов,
полученных в пилотной программе, то в межлабораторной программе должно использоваться не
менее пяти проб. В любом случае необходимо получить не менее 30 степеней свободы как для
повторяемости, так и для воспроизводимости. Для повторяемости это означает получение в итоге не
менее 30 пар результатов в программе.
Для воспроизводимости в Таблице А.1 приведено минимальное число требуемых проб в зависимости
от L, P и Q, где L число лабораторий-участниц, Р и Q отношение оценок составляющих (компонент)
дисперсии, полученных в пилотной программе. При этом Р – отношение компоненты, обусловленной
взаимодействием к компоненте по дублям, а Q – отношение лабораторной компоненты к компоненте
по дублям. В Приложении В приведены формулы, использованные при составлении Таблицы А.1. Если
Q значительно больше Р, то 30-ти степеней свободы для дисперсии получить невозможно; отсутствие
записи в Таблице А.1 соответствует или приближается, к этой ситуации (т.е. требуется более 20 проб).
Для таких случаев вероятны значительные расхождениями между лабораториями.
4.5 Выполнение межлабораторной программы
За выполнение всей программы должен отвечать один человек, от момента распределения документа
на метод испытания и проб до окончательной оценки результатов. Он должен быть знаком с методом
испытания, но не должен сам принимать участие в испытаниях.
Документ с описанием метода испытания должен быть передан всем лабораториям заблаговременно,
чтобы было время на решение возникших вопросов до начала испытаний. Если какая-либо из
лабораторий пожелает попрактиковаться в данном методе испытания заранее, она должна
осуществлять это на других пробах, а не тех, которые будут использовать в программе.
Пробы должны быть складированы, разделены на части и распределены организатором, который
должен хранить резервное количество пробы для непредвиденных случаев. Самое главное, чтобы
6 © ISO 2006 – Все права сохраняются

порции для каждой лаборатории были однородны. Они должны быть закодированы до распределения
и сопровождаться следующей информацией:
а) согласованный проект документа на метод испытания;
b) инструкции по обращению и хранению проб;
с) порядок, в котором пробы необходимо испытывать (для каждой лаборатории случайный порядок
различен);
d) указание о том, что два результата должны быть получены последовательно на каждой
испытуемой пробе одним и тем же оператором на одном и том же оборудовании. По
статистическим причинам крайне важно, чтобы эти два результата получали независимо друг от
друга, т.е. чтобы на второй результат не влияло знание первого результата. Если это считается
невыполнимым для указанного оператора, то пары результатов должны быть получены в режиме
слепого кодирования, но обеспечить получение их в короткий промежуток времени;
е) период времени, в ходе которого необходимо получить повторные результаты, и период времени,
в течение которого все пробы должны быть испытаны;
f) форма протокола для записи результатов. Для каждой пробы должно быть предусмотрено место
для внесения даты выполнения испытания, двух результатов и комментариев по любым
необычным обстоятельствам. Необходимо указать число значащих цифр в отчете о результатах
испытаний;
g) указание, что испытание должно выполняться в нормальных условиях с помощью операторов,
имеющих достаточный опыт, но не обязательно исключительные знания, и чтобы
продолжительность испытания была такой же как обычно.
Операторы, выполнявшие пилотную программу, могут принять участие и в межлабораторной
программе. Если их дополнительный опыт в испытаниях большего (дополнительного) числа проб
оказывает заметное влияние, это должно послужить предупреждением, что метод испытания
неудовлетворительный. Такие операторы должны быть идентифицированы в протоколе испытаний для
того, чтобы любой эффект можно было обнаружить.
5 Исследование результатов межлабораторных испытаний с целью проверки
однородности и выявления выбросов
5.1 Общие положения
В 5.2 – 5.7 установлены процедуры исследования результатов, сообщенных в статистически
обоснованной межлабораторной программе (см. Раздел 4), для выявления следующего:
а) независимость или зависимость показателей прецизионности от уровня результатов испытаний;
b) однородность показателей прецизионности для всех лабораторий;
с) присутствие выбросов аномальных результатов.
Процедуры описаны в математических терминах, основанных на системе обозначений Приложения С и
проиллюстрированных со ссылкой на экспериментальные данные ( пример расчет бромного числа) в
Приложении D.
В 5.2 – 5.7 (и в Разделе 6) используемые процедуры сначала устанавливаются, а затем
иллюстрируются примером с решением на основе данных Приложения D.
В данном разделе предполагается, что все результаты имеют либо простое нормальное
распределение, либо могут быть преобразованы в такое распределение (см. 5.2). Другие случаи
(которые редки) требуют иной обработки, которая выходит за рамки данного международного
стандарта. См. ссылку [8] в отношении статистической оценки на нормальность.
Хотя показанные здесь процедуры даны в форме, удобной для расчетов вручную, настоятельно
рекомендуется использовать электронную вычислительную машину с соответствующей
валидированной программой для хранения и анализа результатов межлабораторного испытания,
основанного на процедурах данного международного стандарта (см., например, Ссылку [9]).
5.2 Преобразование данных
5.2.1 Общие положения
Показатели прецизионности многих методов испытаний зависят от уровня (результатов) испытаний, и,
таким образом, изменчивость сообщенных результатов различается при переходе от пробы к пробе.
Метод исследования выбросов, описанный в данном международном стандарте, требует, чтобы это
было не так, и чтобы такое положение, если необходимо, было исправлено путем преобразования.
Стандартные (среднеквадратические) отклонения лабораторий, D, стандартные (среднеквадратические)
j
отклонения для дублей, d, для j-й пробы рассчитывают и строят на график зависимости их от средних
j
значений по пробе, m. Если через нанесенные таким образом на график точки можно провести две линии,
j
параллельные оси значений т, тогда преобразование не требуется. Однако если через нанесенные точки
можно построить непараллельные оси значений т прямые линии или кривые, которые могут быть описаны
зависимостями D = f (т) и d = f (т), то требуется вводить преобразование.
1 2
Зависимости D = f (т) и d = f (т), в общем, не идентичны. Однако статистические процедуры данного
1 2
международного стандарта требуют, чтобы одно и то же преобразование применялось и для
повторяемости и для воспроизводимости. По этой причине обе зависимости объединяют в одну
зависимость D = f(т) (где D теперь включает и d) путем включения фиктивной переменной, Т. При этом
учитывается различие между зависимостями, если таковое существует, и обеспечивается возможность
выявления данного различия (см. Раздел F.1).
Единую зависимость D = f(т) наилучшим образом оценивают методом взвешенного линейного
регрессионного анализа, даже если в большинстве случаев не взвешенная регрессия дает
удовлетворительное приближение. Вывод весовых функций описан в разделе F.2, и процедура
расчетов для регрессионного анализа описана в разделе F.3. Типичные формы зависимости D = f(т)
приведены в Разделе Е.1. Все они выражены в терминах параметров преобразования В и В .
Оценка В и В и последующая процедура преобразования суммированы в разделе Е.2. Сюда
включены статистические испытания значимости регрессии (т.е. является ли зависимость D = f(т)
параллельной оси т) и значимости различия между зависимостями для повторяемости и
воспроизводимости при 5 %-ном уровне значимости. Если такое различие обнаружено или не
существует подходящее преобразование, тогда следует использовать альтернативные процедуры
“проба за пробой”», описанные в ISO 5725-2. В этом случае невозможно выполнить испытание с целью
выявления систематического смещения лаборатории по всем пробам (см. 5.6) или отдельно оценивать
компоненту дисперсии по взаимодействию (см. 6.2).
Если на 5 %-ном уровне значимости показано, что значимая регрессия формы D = f(т) существует, то
соответствующее преобразование у = F(х), где × – сообщенный результат, задается уравнением:
dx
Fx =K
()

fx
()
(2)
где К – константа. В этом случае все результаты должны быть преобразованы соответствующим
образом, а остальная часть анализа осуществляться в терминах преобразованных результатов.
Типичные преобразования приведены в Раздел Е.1.
8 © ISO 2006 – Все права сохраняются

Сложно сделать выбор преобразования предметом формализованных правил. В конкретных случаях
может потребоваться помощь квалифицированного статистика. Наличие выбросов может повлиять на
правильность решения о выборе типа преобразования (см. 5.7).
5.2.2 Пример с решением
В Таблице 1 перечислены значения т, D и d для восьми проб в примере, приведенном в Приложении
D, округленных до трех значащих цифр. Соответствующие числа степеней свободы даны в скобках.
Таблица 1
Номер
3 8 1 4 5 6 2 7
пробы
m 0,756 1,22 2,15 3,64 10,9 48,2 65,4 114
D
0,066 9 (14) 0,159 (9) 0,729 (8) 0,211 (11) 0,291 (9) 1,50 (9) 2,22 (9) 2,93 (9)
d 0,050 0 (9) 0,057 2 (9) 0,127 (9) 0,116 (9) 0,094 3 (9) 0,527 (9) 0,818 (9) 0,935 (9)
Проверка цифр в Таблице 1 показывает увеличение как D, так и d по мере роста m, причем скоростью
возрастания падает по мере увеличения m. График зависимости этих величин в логарифмических
координатах (то есть график зависимости log D и log d от log m) показывает, что эти точки можно
обоснованно рассматривать как лежащие на двух прямых линиях (см. Рисунок F.1). Из примера с
расчетом, приведенного в Разделе F.4, видно, что градиенты этих линий одинаковы и оцениваются
значением 0,638. Учитывая ошибки в оценке этого значения, для удобства можно применять градиент
равным 2/3.
Следовательно, одно и то же преобразование предназначено для повторяемости и для
воспроизводимости и задается уравнением:
−23 13
x d3xx=

(3)
Поскольку постоянным множителем можно пренебречь, преобразование, таким образом, сводят к
извлечению кубического корня из сообщенных результатов (бромных чисел). Эта процедура дает
преобразованные данные, показанные в Таблице D.2, в которой результат после извлечения
кубического корня приводятся с точностью до третьего знака после запятой.
5.3 Выявление аномальных результатов (выбросов)
5.3.1 Общие положения
Сообщенные данные, или если принято решение о необходимости преобразования, преобразованные
результаты, должны проверяться на аномальные результаты (выбросы). Это такие значения, которые
настолько отличаются от остальных данных, что можно сделать единственное заключение, что эти
данные получились в результате некоторой ошибки при применении метода испытания или испытания
по ошибке другой пробы. Можно использовать различные приемы и менять соответствующие уровни
значимости, но те, которые приведены ниже, считаются подходящими для настоящего международного
стандарта. Эти испытания выбросов основаны на предположении о нормальном распределении
погрешностей (см. 5.1).
5.3.2 Однородность данных, используемых для определения повторяемости
5.3.2.1 Общие положения
Первое испытание для выявления выбросов касается обнаружения противоречивых данных в парах
[1] 2
результатов, полученных при дублировании испытания. Это испытание включает расчет е для всех
ij
комбинаций лаборатория/проба. Далее для испытания отношения наибольшего из значений е к их
ij
сумме используют критерий Кохрена (Cochran) на 1 %-ном уровне значимости (см. Раздел С.5). Если
значение отношения превышает критическое значение, данное в Таблице D.3, соответствующее одной
степени свободы, n – число пар, имеющихся для сравнения, то член пары, наиболее удаленный от
среднего значения по пробе, необходимо отбросить и повторить процесс, сократив n на единицу до тех
пор дальнейшее отбрасывание результатов необходимо. В определенных случаях это испытание
“растет как снежный ком” и приводит к неприемлемо большой доле аномальных результатов (больше
10 %). Если это так, то от такого испытания на выбраковку необходимо отказаться и вернуть несколько
или все отброшенные результаты. В этом случае необходимо принять произвольное решение,
основанное на ситуации.
5.3.2.2 Пример с решением
В случае примера, приведенного в Приложении D, абсолютные разности (расхождения) между
преобразованными результатами повторных испытаний, т.е. парами чисел из Таблицы D.2,
рассчитанными с точностью до третьего десятичного знака, приведены в Таблице 2.
Таблица 2
Образец
Лаборатория
1 2 3 4 5 6 7 8
A 42 21 7 13 7 10 8 0
B 23 12 12 0 7 9 3 0
C 0 6 0 0 7 8 4 0
D 14 6 0 13 0 8 9 32
E 65 4 0 0 14 5 7 28
F 23 20 34 29 20 30 43 0
G 62 4 78 0 0 16 18 56
H 44 20 29 44 0 27 4 32
J 0 59 0 40 0 30 26 0
Самое большое расхождение составляет 0,078 для лаборатории G на пробе 3. Сумма квадратов всех
расхождений равна
22 22
0,042++0,021 K+ 0,026+ 0= 0,043 9

0,078
Таким образом, отношение, которое сравнивают с критерием Кохрена равно = 0,138
0,043 9
В примере используют 72 размаха, а в Таблице D.3 критерий для 80 размахов равен 0,170 9, это
отношение незначимо.
5.3.3 Однородность данных, используемых для определения воспроизводимости
5.3.3.1 Общие положения
Последующие испытания для выявления аномальных результатов относятся к установлению
однородности данных в оценке воспроизводимости и предназначенных для обнаружения либо
аномальной пары результатов одной лаборатории на конкретной пробе, либо аномальной серии
результатов от одной лаборатории по всем пробам. В обоих случаях подойдет испытание по критерию
[2]
Хокинса (Hawkins) .
Процедура включает образование среднего значения по каждой пробе или общего среднего по всем
лабораториям (см. 5.6) отношения наибольшего абсолютного отклонения среднего значения пробы в
10 © ISO 2006 – Все права сохраняются

лаборатории от среднего для пробы (или общего среднего) по всем лабораториям к квадратному
корню из соответствующих сумм квадратов (см. Раздел С.6).
Отношение, соответствующее наибольшему абсолютному отклонению, следует сравнивать с
критическим значением на 1 %-ном уровне значимости, приведенным в Таблице D.4, где n число
ячеек-комбинаций лаборатория/проба для рассматриваемой пробы (или число средних по всем
лабораториям), и где v число степеней свободы для суммы квадратов, которая является дополнением
к квадрату, соответствующему рассматриваемой пробе. В испытании ячеек лаборатория/проба v
относится к остальным пробам, однако в испытаниях средних значений по всем лабораториям v будет
равно нулю.
Если значимый результат получают для отдельных проб, то соответствующие экстремальные
значения должны быть отброшены и процедуру испытания повторяют. Если экстремальные значения
обнаружены среди итогов по лабораториям, то все результаты, полученные данной лабораторией,
должны быть отброшены.
Если анализ методом “снежного кома” ведет к неприемлемо большой доле выбросов (скажем, больше
10 %), то от такого браковочного испытания необходимо отказаться, и некоторые или все аномальные
результаты следует вернуть. Необходимо принять произвольное решение, основанное на ситуации.
5.3.3.2 Пример с решением
Применение испытания по критерию Хокинса к средним значениям по ячейкам для пробы показано
ниже.
Первый шаг состоит в том, чтобы рассчитать отклонение средних значений по ячейкам от
соответствующих средних значений по пробам для всего массива. Эти значения показаны в Таблице 3
с точностью до третьего десятичного разряда.
Затем для каждой пробы рассчитывают сумму квадратов отклонений. Эти результаты показаны в
Таблице 3 также с точностью до единиц в третьем десятичном разряде.
Таблица 3
Проба
Лаборатория
1 2 3 4 5 6 7 8
A
20 8 14 15 10 48 6 3
B
75 7 20 9 10 47 6 3
C
64 35 3 20 30 4 22 25
D
314 33 18 42 7 39 80 50
E
32 32 30 9 7 18 18 39
F
75 97 31 20 30 8 74 53
G
10 34 32 20 20 61 9 62
H
42 13 4 42 13 21 8 50
J
1 28 22 29 14 8 10 53
Сумма
квадратов 117 15 4 6 3 11 13 17
В первую очередь испытывают ячейку с наибольшим экстремальным отклонением. Это отклонение
получено лабораторией D по пробе 1. Соответствующее значение отношения в испытании по критерию
Хокинса равно
0,314
B*0==,7281
0,117++0,015 .+ 0,017
Критическое значение, соответствующее n = 9 ячейкам для пробы 1 и v = 56 степеням свободы для
других проб, получают путем интерполяции значений из Таблицы D.4 равным 0,372 9. Значение,
полученное при испытании, больше критического значения, и поэтому результаты испытания,
полученные в лаборатории D на образце 1, отбрасывают.
Так как был отброшен аномальный результат, среднее значение, отклонения и сумму квадратов для
пробы 1 пересчитывают, а процедуру выявления аномальных результатов повторяют. Следующая
ячейка для испытания – это ячейка с результатом, полученным лабораторией F по пробе 2.
Соответствующее значение отношения В при испытании по Хокинсу будет
0,097
B*0==,3542
0,006++0,015 .+ 0,017
Критическое значение, соответствующее n = 9 ячейкам для пробы 2 и v = 55 степеням свободы
соответствующим алгебраическому дополнению, получают интерполяцией значений из Таблицы D.4
равным 0,375 6. Поскольку экспериментальное отношение меньше чем критическое значение, никакие
результаты не отбрасывают.
5.4 Браковка полного набора данных по пробе
5.4.1 Общие положения
Межлабораторное среднеквадратическое отклонение и среднеквадратическое отклонение для
повторных испытаний необходимо проанализировать на наличие выпадающих проб. Если уже
выполнялось какое-либо преобразование или какие-либо результаты были отброшены как аномальные,
необходимо вычислить новые значения среднеквадратических отклонений.
Если среднеквадратическое отклонение для какой-либо пробы слишком велико, необходимо
произвести исследование с точки зрения исключения всех результатов испытания этой пробы.
Можно использовать критерий Кохрена на 1 %-ном уровне значимости, если среднеквадратические
отклонения основаны на одинаковом числе степеней свободы. Это включает расчет отношения
наибольшей из соответствующих сумм квадратов (межлабораторные или для повторных испытаний в
зависимости от ситуации) к их общей сумме (см. Раздел С.5). Если это отношение превышает
критическое значение, приведенное в Таблице D.3 с n в качестве числа проб, с v в качестве числа
степеней свободы для отдельного среднеквадратического отклонения, то все результаты для
рассматриваемой пробы необходимо отбросить. В этом случае необходимо проверить, не
обусловлено ли экстремальное стандартное отклонение применением неподходящего преобразования
(см. 5.2) или необнаруженными аномальными результатами.
Не существует оптимального критерия для исследования однородности стандартных отклонений,
основанных на различных числах степеней свободы. Однако распределение отношения максимальной
дисперсии к дисперсии средневзвешенной по остальным пробам подчиняется F-распределению с v и
v с
...

Questions, Comments and Discussion

Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.

Loading comments...