ISO/DIS 22110

Milk and milk products -- Sampling -- Inspection by variables in the presence of measurement error

Lait et produits laitiers -- Échantillonnage -- Contrôle par variables en présence d'erreur de mesure

Standards Content (sample)

DRAFT INTERNATIONAL STANDARD ISO/DIS 22110
DRAFT INTERNATIONAL STANDARD ISO/IDF 207
ISO/TC 34/SC 5 Secretariat: NEN
Voting begins on: Voting terminates on:
2005-05-27 2005-10-27

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION • МЕЖДУНАРОДНАЯ ОРГАНИЗАЦИЯ ПО СТАНДАРТИЗАЦИИ • ORGANISATION INTERNATIONALE DE NORMALISATION

Milk and milk products — Sampling — Inspection by variables in
the presence of measurement error

Lait et produits laitiers — Échantillonnage — Contrôle par variables en présence d'erreur de mesure

ICS 67.100.01

In accordance with the provisions of Council Resolution 15/1993 this document is circulated in

the English language only.

Conformément aux dispositions de la Résolution du Conseil 15/1993, ce document est distribué

en version anglaise seulement.
This draft is submitted to a parallel enquiry vote in ISO and IDF.

To expedite distribution, this document is circulated as received from the committee secretariat.

ISO Central Secretariat work of editing and text composition will be undertaken at publication

stage.

Pour accélérer la distribution, le présent document est distribué tel qu'il est parvenu du

secrétariat du comité. Le travail de rédaction et de composition de texte sera effectué au

Secrétariat central de l'ISO au stade de publication.

THIS DOCUMENT IS A DRAFT CIRCULATED FOR COMMENT AND APPROVAL. IT IS THEREFORE SUBJECT TO CHANGE AND MAY NOT BE

REFERRED TO AS AN INTERNATIONAL STANDARD UNTIL PUBLISHED AS SUCH.

IN ADDITION TO THEIR EVALUATION AS BEING ACCEPTABLE FOR INDUSTRIAL, TECHNOLOGICAL, COMMERCIAL AND USER PURPOSES, DRAFT

INTERNATIONAL STANDARDS MAY ON OCCASION HAVE TO BE CONSIDERED IN THE LIGHT OF THEIR POTENTIAL TO BECOME STANDARDS TO

WHICH REFERENCE MAY BE MADE IN NATIONAL REGULATIONS.

© International Organization for Standardization and International Dairy Federation, 2005

---------------------- Page: 1 ----------------------
ISO/DIS 22110ΩΩΩΩ IDF 207
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.

Copyright notice

This ISOIDF document is a Draft International Standard and is copyright-protected by ISO and IDF. Except

as permitted under the applicable laws of the user’s country, neither this ISOIDF draft nor any extract from

it may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic,

photocopying, recording or otherwise, without prior written permission being secured.

Requests for permission to reproduce should be addressed to ISO or IDF at the respective address below.

ISO copyright office International Dairy Federation

Case postale 56  CH-1211 Geneva 20 Diamant Building  Bd Auguste Reyers 80  B-1030 Brussels

Tel. + 41 22 749 01 11 Tel. + 32 2 733 98 88
Fax + 41 22 749 09 47 Fax + 32 2 733 04 13
E-mail [e-mail] E-mail [e-mail]
Web www.iso.org Web www.fil-idf.org
Reproduction may be subject to royalty payments or a licensing agreement.
Violators may be prosecuted.
ii ISO and IDF 2005 – All rights reserved
---------------------- Page: 2 ----------------------
ISO/DIS 22110 | IDF 207
Contents Page

Foreword ............................................................................................................................................................iv

Foreword .............................................................................................................................................................v

1 Scope......................................................................................................................................................6

2 Normative references............................................................................................................................6

3 Terms and definitions ...........................................................................................................................7

4 Types of inspection plans ....................................................................................................................7

5 Assessment of statistical risk and confidence ..................................................................................8

6 Classification of defects .....................................................................................................................10

6.1 Choice of inspection and acceptable quality level ..........................................................................11

7 Selection of sampling plans...............................................................................................................11

8 Sampling and inspection procedures ...............................................................................................13

9 Verification in the presence of non-negligible between-laboratory measurement

uncertainty ...........................................................................................................................................15

9.1 Verification procedure........................................................................................................................16

9.2 Verification of inspection by Variables outcome .............................................................................16

9.3 Verification of a long term process standard deviation..................................................................18

9.4 Verification of a lot-based process standard deviation ..................................................................18

Annex A .............................................................................................................................................................20

Annex B ............................................................................................................................................................21

Annex C .............................................................................................................................................................27

Annex D .............................................................................................................................................................29

Annex E .............................................................................................................................................................35

Annex F..............................................................................................................................................................37

Annex G.............................................................................................................................................................42

Bibliography......................................................................................................................................................43

© ISO 2004 – All rights reserved iii
DRAFT 2005
---------------------- Page: 3 ----------------------
ISO/DIS 22110 | IDF 207
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 22110 | IDF 207 was prepared by Technical Committee ISO/TC 34, Food, Subcommittee SC 5, Milk and

milk products and the International Dairy Federation (IDF). It is being published jointly by ISO and IDF.

This first edition of ISO 22110 | IDF 207 cancels and replaces the first edition of ISO 8197:1988, which is fully

covered by this document.
© ISO 2004 – All rights reserved iv
DRAFT 2005
---------------------- Page: 4 ----------------------
ISO/DIS 22110 | IDF 207
Foreword

IDF (the International Dairy Federation) is a worldwide federation of the dairy sector with a National

Committee in every member country. Every National Committee has the right to be represented on the IDF

Standing Committees carrying out the technical work. IDF collaborates with ISO and AOAC International in

the development of standard methods of analysis and sampling for milk and milk products.

Draft International Standards adopted by the Action Teams and Standing Committees are circulated to the

National Committees for voting. Publication as an International Standard requires approval by at least 50% of

IDF National Committees casting a vote.

Attention is drawn to the possibility that some of the elements of this document may be the subject of patent

rights. IDF shall not be held responsible for identifying any or all such patent rights.

ISO 22110 | IDF 207 was prepared by Technical Committee ISO/TC 34, Food products, Subcommittee SC 5,

Milk and milk products and the International Dairy Federation (IDF). It is being published jointly by ISO and

IDF.

All work was carried out by the Joint ISO/IDF Action Team on Statistics of analytical data, of the Standing

Committee on Quality assurance, statistics of analytical data and sampling, under the aegis of its project

leader, Dr. R. Crawford (NZ).

This first edition of ISO 22110 | IDF 207 cancels and replaces the first edition of IDF 136A:1992, which is fully

covered by this document.
© ISO 2004 – All rights reserved v
DRAFT 2005
---------------------- Page: 5 ----------------------
ISO/DIS 22110 | IDF 207
Milk and milk products — Sampling — Inspection by variables
in the presence of measurement error
1 Scope

This standard provides guidance in the selection of appropriate acceptance sampling schemes for the

inspection of dairy products submitted for inspection in lots.

There are many situations in which product is transferred from one organisation to another. In these situations,

either or both organisations may use sampling plans to satisfy themselves that the product is of acceptable

quality.

Because testing of milk and milk products is usually destructive and costly, it is not feasible to test all product

produced. Instead, samples are taken and tested. There is always uncertainty involved with sampling. In the

case of acceptance inspection plans this means that there is a risk that product could be incorrectly classified,

with downstream costs. This risk is not unique to milk and milk products, but is somewhat more noticeable

than in some other commodities because of the inherent variability of milk and milk products, uncertainties in

the test methods and the costs of testing.

The sampling plans in this standard are intended for using by receivers under the following conditions:

a) The inspection procedure is to be applied to a lot of product, supplied by one producer using one

production process. Where product is transferred in consignments consisting of collections of lots,

assessments should be made on an individual lot basis.

b) There is only a single quality characteristic under consideration. The plans can be applied to more than

one characteristic but the individual risk profiles shall be adjusted to control the overall risk or items shall

be classified as pass/fail and an attribute inspection plan used. It is recommended that a statistician be

consulted.
c) The production process is stable.

d) The quality characteristic follows, or can be transformed to, a normal distribution.

e) The measurement process is stable and the measurement errors are normally distributed.

The lot of product may be packaged into discrete units, be unpackaged or be bulk packaged.

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 2859, Sampling procedures for inspection by attributes.

ISO 3951:1989, Sampling procedures and charts for inspection by variables for percent nonconforming.

ISO 3534-1:1993, Statistics — Vocabulary and symbols. Part 1. Probability and general statistical terms.

ISO 5725 (all parts), Accuracy (trueness and precision) of measurement methods and results.

© ISO 2004 – All rights reserved 6
DRAFT 2005
---------------------- Page: 6 ----------------------
ISO/DIS 22110 | IDF 207

ISO/TR 8550, Guide for the selection of an acceptance sampling system, scheme or plan for inspection of

discrete items in lots.
ISO 707 (IDF 50), Milk and milk products — Guidance on sampling.

ISO 5538 | IDF 113, Milk and milk products — Sampling — Inspection by attributes.

ISO 8197 (IDF 136), Milk and milk products — Sampling — Inspection by variables.

ISO 17025, General requirements for the competence of testing and calibration laboratories

3 Terms and definitions

For the purposes of this standard, the terms and definitions given in ISO 3534 and the following apply.

3.1
continuous result

result from a method that is a decimal number, e.g. the fat content of milk powder as measured by the Röse-

Gottlieb method.
3.2
nominal result

result from a method that delivers a finite number of outcomes, such as pass/fail, present/absent, a number or

letter on a limited scale, e.g. {A, B, C, D, E} or {1, 2, 3, 4}, or a colour on a colour chart.

4 Types of inspection plans

An acceptance-sampling plan is intended for determining the acceptance or rejection of a lot (see ISO 3534).

The plan stipulates the number of units (items) in the sample, to be drawn randomly from a lot for inspection

against the product specification. The lot is then classified as "acceptable" or "not acceptable" according to

how the inspection results compare with the criteria of the acceptance-sampling plan.

There are two main categories of acceptance inspection schemes: inspection by variables sampling plans and

inspection by attribute sampling plans. These differ primarily in the type of quality characteristic under

assessment: continuous or nominal. This standard considers only the former, although, for completeness, this

section, Section 4, describes inspection by attribute plans in general terms.
4.1 Inspection by variables schemes

Inspection by variables are methods which consists in measuring a quantitative characteristic of each item of

a population or of a sample taken from this population (see ISO 3534). In practice these sampling schemes

use measurements made on each sample to determine the acceptability of each lot of product. In an

inspection by variables scheme, the sample average and the standard deviation are calculated. These are

then used to assess the acceptability of the lot.

Most inspections by variables schemes are based on the assumption that the data are normally distributed or

can be transformed so that they are normally distributed. Such sampling plans are described in detail in ISO

3951 when the test measurement uncertainty is negligible. However, ISO 3951 does not consider situations in

which there is non-negligible measurement error. This standard expands upon the methods used in ISO 3951

to encompass situations where there is non-negligible measurement error.
© ISO 2004 – All rights reserved 7
DRAFT 2005
---------------------- Page: 7 ----------------------
ISO/DIS 22110 | IDF 207
4.2 Inspection by attribute schemes

Inspection by attributes are methods which consists in taking note, for every item of a population or of a

sample taken from this population, of the presence or absence of a certain characteristic (attribute) and in

counting how many items have or do not have this characteristic (ISO 3534).

These schemes categorises each sampled item as either acceptable or not acceptable. The number of items

in each category is used to assess the overall acceptability of each lot of product. However, note that nominal

methods may also be subject to measurement error; consideration of these is outside the scope of this

standard. Inspection by attribute sampling plans does not make assumptions about the distribution of the data

and are therefore completely general. However, the cost of being completely general is that larger sample

sizes are typically required to give the same level of confidence as would be obtained from the equivalent

inspection by variables scheme. Suitable inspection by attribute plans is described in ISO 2859 and

ISO 5538 | IDF 113.
5 Assessment of statistical risk and confidence
5.1 General

Sampling plans are designed to separate, with confidence, lots that contain acceptably low levels of non-

conforming product from lots that contain unacceptably high levels of non-conforming product. Lots that

contain very little non-conforming product should be accepted with high probability and lots with a high

proportion of non-conforming product should have a low probability of being accepted.

For all acceptance-sampling schemes, it is possible to calculate the probability of accepting a lot based on the

proportion of the lot that is non-conforming. A graph showing the probability of acceptance against the

proportion of non-conforming product is called an operating characteristic (OC) curve. These OC curves

summarise the statistical risk and confidence of the acceptance inspection scheme. A typical OC curve is

shown in the figure 1.
100%
80%
60%
40%
20%
0% 20% 40% 60% 80% 100%
% non-co nform ing in th e lo t
Figure 1 — Example of a typical OC curve

There are an infinite number of OC curves that can have a gradual slope or a steep slope. The overall shape

of the OC curve determines the power of the acceptance inspection scheme to differentiate between lots of

product of acceptable quality and lots of product that are not of acceptable quality.

5.2 Descriptors of OC curves

Two main descriptors are used to index OC curves: the acceptable quality level and the limiting quality.

© ISO 2004 – All rights reserved 8
DRAFT 2005
Prob accept
---------------------- Page: 8 ----------------------
ISO/DIS 22110 | IDF 207
5.2.1 Acceptable quality level

The Acceptable Quality Level (AQL) a quality level which in a sampling plan corresponds to a specified but

relatively high probability of acceptance (see ISO 3534). In practice this is the maximum percentage or

proportion of nonconformities in a lot or batch that can be considered satisfactory as a process average.

Three OC curves with the same AQL are shown in the figure below. All three curves pass through the one

point — (AQL with 10 % non-conforming product, 90 % probability of acceptance) — but they differ in their

steepness. For a fixed AQL, acceptance inspection schemes using larger sample sizes have steeper OC

curves. That is, as the sample size increases, the plans are better able to differentiate between product of

acceptable quality (less than AQL non-conforming) and marginal product.
Operating Characteristic Curves Based of Acceptable Quality Level
(AQL= 10%)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% non-conforming in the lot
Figure 2 — Operating Characteristic Curves based of Acceptable Quality Level
(AQL= 10 %)
5.2.2 Limiting quality

The Limiting Quality (LQ) in a sampling plan is a quality level which corresponds to a specified and relatively

low probability of acceptance (see ISO 3534). In practice the LQ is the quality level, when a lot is considered

in isolation, which, for the purposes of acceptance sampling inspection, is limited to a low probability of

acceptance. Quantitatively it is the percentage of non-conforming product for which there is approximately

10 % chance of acceptance.

The LQ defines the poorest level of marginal quality. LQ is sometimes referred to as lot tolerance percent

defective (LTPD) and rejectable quality level (RQL). Four OC curves with the same LQ are shown in the figure

below. All four OC curves pass through the one point (LQ 30 % non-conforming product, 90 % probability of

rejection) but they differ in their steepness.

For a fixed LQ, acceptance inspection schemes with larger sample sizes have steeper OC curves. That is, as

the sample size increases, the plans are better able to differentiate between marginal product (less than LQ

non-conforming) and unacceptable product.
© ISO 2004 – All rights reserved 9
DRAFT 2005
Probability of Acceptance
---------------------- Page: 9 ----------------------
ISO/DIS 22110 | IDF 207
Operating Characteristic Curves Based on Limiting Quality Level
(LQ= 30%)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
% non-conforming in the lot
Figure 3 — Operating Characteristic Curves Based on Limiting Quality Level
(LQ= 30 %)
5.2.3 Sample size

Traditionally, inspection-sampling plans have been applied only to lots consisting of discrete units. However,

this constraint is unnecessary — inspection by variables sampling plans also apply to continuous product. In

addition, previous standards have also tended to vary the sample size according to the lot size even though

the sample size is, in the strict statistical sense, independent of the lot size.

The designers of the plans, however, such as those in ISO 3951, have deliberately chosen to relate sample

size to lot size to force steeper OC curves upon larger lots, to reduce the chance of making an incorrect

decision about larger lots. This approach is generally not appropriate for milk and milk products because, in

many situations, the size of a packaged unit is arbitrary and not related to the quantity of product

manufactured. For example, milk powder is usually packaged in either 25 kg bags or 900 kg bulk bins, but the

customer typically purchases product by either the kilogram or the tonne, with neither basis related to the

quantity of product manufactured. Relating the sample size to the number of units in the lot (as per ISO 3951)

could produce overall risk for the same amount of product depending on how product units are defined.

This standard takes a different approach to the determination of sample size from that used in ISO 3951. This

approach is fully described in annex D.
6 Classification of defects

The contract or specification shall clearly define and document all critical, major and minor defects in an

unambiguous way, and define the inspection sampling plans. The following terms and definitions on defects

applies:
6.1
critical defect
defect that is likely to make the product unacceptable.
6.2
major defect
defect that is likely to make the product unfit for sale to a consumer.

EXAMPLE A major defect could result in spoilage or contamination with an inhibitory substance.

© ISO 2004 – All rights reserved 10
DRAFT 2005
Probability of Acceptance
---------------------- Page: 10 ----------------------
ISO/DIS 22110 | IDF 207
6.3
minor defect

failure to comply with a specification, but does not make the unit unfit for use or cause it to spoil.

EXAMPLE The chemical composition or net content falling outside, but close to, a specification limit would usually

be considered to be a minor defect.
6.1 Choice of inspection and acceptable quality level

1. Sampling plans for major defects shall be selected using an AQL of not more than 6,5 %.

2. Sampling plans for minor defects shall be selected using an AQL of not more than 10 %.

3. The risk profile (B–N) is selected to control the risk of accepting a lot when more than the AQL is non-

conforming.

4. Risk category B plans have the highest risk of wrongly accepting a lot when more than proportion

AQL is non-conforming. These plans are more suitable for small lots.

5. Risk category N plans have the lowest risk of wrongly accepting a lot when more than proportion AQL

is non-conforming.
7 Selection of sampling plans

The flow chart in figure 4 describes the process for selecting a sampling plan for a particular application from

the options presented in this standard.
The following outlines the basic procedure.
a) Determine the appropriate AQL using the guidelines in clause 6.
b) Select an appropriate risk profile for the application — see clause 6.

c) Determine the type of result associated with the quality parameter under consideration; that is, are the

data continuous or nominal?

d) If the data are nominal, then an attributes sampling plan is required. Select an appropriate plan from

ISO 2859 or ISO 5538 | IDF 113 or use a three-class sampling plan.

e) If the data are continuous but neither normally distributed nor transformable to normality, then results

should be coded as acceptable/not acceptable with respect to some threshold value, and an attributes

sampling plan from ISO 2859 or ISO 5538 | IDF 113 should be used.

f) If the data are continuous and are normally distributed, or can be transformed to normality, then the

following should be used.

1. If the measurement error is negligible, then the ISO 3951 standard may be used. Equivalently,

depending on whether the process standard deviation is known or unknown, the tables in annexes A

and B, respectively, can be used with γ = 0. Statistical details are explained in annexes E and F.

2. If the between-laboratory measurement error is not negligible, then there is no valid inspection by

variables sampling plans for isolated lots and a verification scheme must be used instead. Refer to

Section 9 and annex 9.

3. If the between-laboratory measurement error is negligible but there is non-negligible (known) within-

laboratory measurement error (as measured by the repeatability standard deviation), and if the

process standard deviation is known, then the tables in annex A (known process standard deviation)

should be used to determine the sampling plan, i.e. the sample size and the acceptability constant.

© ISO 2004 – All rights reserved 11
DRAFT 2005
---------------------- Page: 11 ----------------------
ISO/DIS 22110 | IDF 207

4. If the between-laboratory measurement error is negligible but there is non-negligible within-laboratory

measurement error (as measured by the repeatability standard deviation), and if the process standard

deviation is unknown but the ratio γ of the within-laboratory standard deviation to the process standard

deviation is known, then the tables in annex B (unknown process standard deviation, known ratio)

should be used to determine the sampling plan. Note that the ratio of the repeatability standard

deviation to the process standard deviation, γ, need be known only approximately.

5. If the between-laboratory measurement error is negligible but there is non-negligible within-laboratory

measurement error (as measured by the repeatability standard deviation), and if the process standard

deviation is unknown, then the tables in annex C (unknown process standard deviation, known within-

laboratory measurement error) should be used to determine the sampling plan.
Determ ine
AQ L
Determ ine
R isk L eve l
Yes
N o m inal D ata ?
T ransform a ble to U se In sp ection b y
N orm a l D a ta ?
Norm al? Attributes
Yes
Yes
N on-N egligible
U se IS O 3 951
m easurem ent E rror?
Yes
S ign ificant B e tw een No
U se V erifica io tn S ch e m e
Lab E rro r?
Yes
Yes
P roc ess S td de v In sp e ctio n b y V a ria b le s:
Known? Ann e x A
Yes
In sp e ctio n by va ria b le s
Ratio γ kn ow n?
Ane x B
In spe ction b y va riab le s -
A nnex C
Figure 4 — Flow chart
© ISO 2004 – All rights reserved 12
DRAFT 2005
---------------------- Page: 12 ----------------------
ISO/DIS 22110 | IDF 207

NOTE The measurement error can be considered to be non-negligible if the total measurement error standard

deviation exceeds 30 % of the process standard deviation. The total measurement error standard deviation includes both

within-laboratory and between-laboratory variation.

It is possible to reduce the effective measurement uncertainty by performing analyses in replicate and utilising

the average result for each physical sample because the measurement error standard deviation of the

average is, by the central limit theorem, σ / n where σ is the measurement error standard deviation of

r 2 r
individual results and n is t
...