Standard ISO TR 22514 4 2007 9658 PDF

TECHNICAL ISO/TR
REPORT 22514-4
First edition
2007-12-01
Statistical methods in process

management — Capability and
performance —
Part 4:
Process capability estimates and
performance measures
Méthodes statistiques dans la gestion de processus — Aptitude et
performance —
Partie 4: Estimations de l'aptitude de processus et mesures de
performance
--`,,```,,,,````-`-`,,`,,`,`,,`---
Reference number
ISO/TR 22514-4:2007(E)
Copyright International Organization for Standardization © ISO 2007

Provided by IHS under license with ISO
No reproduction or networking permitted without license from IHS Not for Resale
ISO/TR 22514-4:2007(E)
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 PROTECTED DOCUMENT

© ISO 2007
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
Copyright International Organization for Standardization © ISO 2007 – All rights reserved
ISO/TR 22514-4:2007(E)
Contents Page
Foreword............................................................................................................................................................ iv
Introduction ........................................................................................................................................................ v
1 Scope ......................................................................................................................................................1
2 Terms and definitions ...........................................................................................................................1
2.1 Basic terms.............................................................................................................................................1
2.2 Process capability, estimates and indices..........................................................................................5
2.3 Process performance, measures and indices ....................................................................................8
3 Symbols and abbreviated terms ........................................................................................................10
3.1 Symbols ................................................................................................................................................10
3.2 Abbreviated terms ...............................................................................................................................12
4 Statistical measures used in process capability and performance ...............................................13
4.1 General..................................................................................................................................................13
4.2 Measures of location ...........................................................................................................................13
4.3 Measures of dispersion.......................................................................................................................13
4.4 Mean square error (MSE) ....................................................................................................................13
4.5 Reference limits ...................................................................................................................................13
4.6 Reference interval, also known as process spread .........................................................................13
4.7 Attributes ..............................................................................................................................................13
5 Capability..............................................................................................................................................14
5.1 General..................................................................................................................................................14
5.2 Process capability ...............................................................................................................................15
5.3 Process location ..................................................................................................................................16
5.4 Process capability indices for measured data .................................................................................17
5.5 Process capability indices for measured data (non-normal) ..........................................................19
5.6 Alternative method for describing and calculating process capability estimates .......................21
5.7 Other capability indices for measured data in other special cases ...............................................22
5.8 Assessment of proportion out-of-specification (normal distribution) ...........................................23
5.9 Attributes ..............................................................................................................................................24
6 Performance .........................................................................................................................................25
6.1 General..................................................................................................................................................25
6.2 Process performance indices for measured data (normal distribution)........................................26
6.3 Process performance indices for measured data (non-normal distribution) ................................27
6.4 Other performance indices for measured data.................................................................................28
6.5 Assessment of proportion out-of-specification (normal distribution) ...........................................28
6.6 Attributes ..............................................................................................................................................28
Annex A (informative) Estimating standard deviations.................................................................................29
Annex B (informative) Estimating capability and performance measures using Pearson curves —
Procedure and example ......................................................................................................................31
Annex C (informative) Distribution identification ..........................................................................................39
--`,,```,,,,````-`-`,,`,,`,`,,`---
Annex D (informative) Confidence intervals...................................................................................................44

Annex E (informative) Example of the output from computer software ......................................................46
Bibliography ......................................................................................................................................................48
© ISO 2007 – All rights reserved

Copyright International Organization for Standardization iii
ISO/TR 22514-4:2007(E)
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.
In exceptional circumstances, when a technical committee has collected data of a different kind from that
which is normally published as an International Standard (“state of the art”, for example), it may decide by a
simple majority vote of its participating members to publish a Technical Report. A Technical Report is entirely
informative in nature and does not have to be reviewed until the data it provides are considered to be no
longer valid or useful.
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/TR 22514-4 has been prepared by Technical Committee ISO/TC 69, Applications of statistical methods,
Subcommittee SC 4, Applications of statistical methods in process management.
ISO 22514 consists of the following parts, under the general title Statistical methods in process management
— Capability and performance:
⎯ Part 1: General principles and concepts
⎯ Part 3: Machine performance studies for measured data on discrete parts
⎯ Part 4: Process capability estimates and performance measures [Technical Report]
In the future, it is planned to revise ISO 21747:2006 (Statistical methods — Process performance and
capability statistics for measured quality characteristics) as Part 2.
NOTE ISO/TR 22514-4 was initially prepared as ISO/DTR 12783. It was renumbered before publication to include it
in the ISO 22514 series.
--`,,```,,,,````-`-`,,`,,`,`,,`---
iv
ISO/TR 22514-4:2007(E)
Introduction
Many organizations have embarked upon a continuous improvement strategy. To comply with such a strategy,
any organization will need to evaluate the capability and performance of its key processes. The methods
described in this part of ISO 22514 are intended to assist any management in this respect. These evaluations
need to be constantly reviewed by management so that actions compatible with continuous improvement can
be taken when required.
The content of this part of ISO 22514 has been subject to large shifts of opinion during recent times and this
report attempts to reflect the current position. The most fundamental shift has been to philosophically separate
what is named in this document as capability conditions from performance conditions, the primary difference
being whether statistical stability has been obtained (capability) or not (performance). This naturally leads onto
the two sets of indices that are to be found in their relevant clauses. It has become necessary to draw a firm
distinction between these since it has been observed in industry that companies have been deceived about
their true capability position due to inappropriate indices being calculated and published.
The progression of this part of ISO 22514 is from the general condition to the specific and this approach leads
to general formulae being presented before their more usual but specific manifestations.
--`,,```,,,,````-`-`,,`,,`,`,,`---
There exist numerous references that describe the importance of understanding the processes at work within
any organization, be it a manufacturing process or an information handling process. As organizations compete
for sales with each other, it has become increasingly apparent that it is not only the price paid for a product or
service that matters so much, but also what costs will be incurred by the purchaser from using such a product
or service. The objective for any supplier is to continually reduce variability and not to just satisfy specification.
Continual improvement leads to reductions in the costs of failure and assists in the drive for survival in an
increasingly more competitive world. There will also be savings in appraisal costs for as variation is reduced
the need to inspect product might disappear or the frequency of sampling might be reduced.
Process capability and performance evaluations are necessary to enable organizations to assess the
capability and performance of their suppliers. Those organizations will find the indices contained within this
part of ISO 22514 useful in this endeavour.
Quantifying the variation present within a process enables judgement of its suitability and ability to meet some
given requirement. The following paragraphs and clauses provide an outline of the philosophy required to be
understood to determine the capability or performance of any process.
All processes will be subject to certain inherent variability. This part of ISO 22514 does not attempt to explain
what is meant by inherent variation, why it exists, where it comes from nor how it affects a process. This part
of ISO 22514 starts from the premise that it exists and is stable.
Process owners should endeavour to understand the sources of variation in their processes. Methods such as
flowcharting the process and identifying the inputs and outputs from a process assist in identification of these
variations together with the appropriate use of cause and effect (fishbone) diagrams.
It is important for the user of this part of ISO 22514 to appreciate that variations exist that will be of a short-
term nature as well as those that will be of a long-term nature and that capability determinations using only the
short-term variation might be greatly different to those which have used the long-term variability.
When considering short-term variation, a study that uses only the shortest-term variation, sometimes known
as a machine study, might be carried out. The method required to carry out such a study will be outside the
scope of this part of ISO 22514; however, it should be noted that such studies are important and useful.
It should be noted that where the capability indices given in this part of ISO 22514 are computed, they only
form point estimates of their true values. It is therefore recommended that, wherever possible, the indices'
confidence intervals are computed and reported. This part of ISO 22514 describes methods by which these
can be computed.

Copyright International Organization for Standardization v
--`,,```,,,,````-`-`,,`,,`,`,,`---
Copyright International Organization for Standardization

TECHNICAL REPORT ISO/TR 22514-4:2007(E)
Statistical methods in process management — Capability and

performance —
Part 4:
Process capability estimates and performance measures
1 Scope
This part of ISO 22514 describes process capability and performance measures that are commonly used.
2 Terms and definitions

For the purposes of this document, the following terms and definitions apply.
2.1 Basic terms
2.1.1
product
result of a process
NOTE Process is defined in ISO 12207:1995, 3.17 and in ISO 9000:2005, 3.4.1.
2.1.2
characteristic
distinguishing feature
NOTE 1 A characteristic can be inherent or assigned.
NOTE 2 A characteristic can be qualitative or quantitative.
NOTE 3 There are various classes of characteristics, such as the following:
⎯ physical (e.g. mechanical, electrical, chemical or biological characteristics);
⎯ sensory (e.g. relating to smell, touch, taste, sight, hearing);
⎯ behavioural (e.g. courtesy, honesty, veracity);
⎯ temporal (e.g. punctuality, reliability, availability);
⎯ ergonomic (e.g. physiological characteristic, or related to human safety);
⎯ functional (e.g. maximum speed of an aircraft).
[ISO 9000:2005, 3.5.1]
--`,,```,,,,````-`-`,,`,,`,`,,`---

Copyright International Organization for Standardization 1
ISO/TR 22514-4:2007(E)
2.1.3
quality characteristic
inherent characteristic (2.1.2) of a product (2.1.1), process or system related to a requirement
NOTE 1 Inherent means existing in something, especially as a permanent characteristic.
NOTE 2 A characteristic assigned to a product, process or system (e.g. the price of a product, the owner of a product)
is not a quality characteristic of that product, process or system.
[ISO 9000:2005, 3.5.2]
2.1.4
specification limit
limiting value stated for a characteristic (2.1.2)
[ISO 3534-2:2006, 3.1.3]
2.1.5
specified tolerance
difference between the upper specification limits and lower specification limits (2.1.4)
[ISO 3534-2:2006, 3.1.6]

--`,,```,,,,````-`-`,,`,,`,`,,`---
2.1.6
target value
T
preferred or reference value of a characteristic (2.1.2) stated in a specification
[ISO 3534-2:2006, 3.1.2]
2.1.7
distribution
〈of a characteristic〉 information on the probabilistic behaviour of a characteristic (2.1.2)
NOTE 1 The distribution of a characteristic (2.1.2) can be represented, for example, by ranking of the values of the
characteristic (2.1.2) and showing the resulting pattern of measures or scores in the form of a tally chart or histogram.
Such a pattern provides all of the numerical value information of the characteristic (2.1.2) except for the serial order in
which the data arises.
NOTE 2 The distribution of a characteristic (2.1.2) is dependent on prevailing conditions. Thus, if meaningful
information about the distribution of a characteristic (2.1.2) is desired the conditions under which the data are collected
should be specified.
NOTE 3 It is important to know the class of distributions (2.1.8), for instance, normal or log-normal, before predicting
or estimating process capability or performance measures and indices or fraction nonconforming.
[ISO 3534-2:2006, 2.5.1]
2.1.8
class of distributions
particular family of distributions (2.1.7) each member of which has the same common attributes by which the
family is fully specified
EXAMPLE 1 The two-parameter, symmetrical bell-shaped, normal distribution with parameters mean and standard
deviation.
EXAMPLE 2 The three-parameter Weibull distribution with parameters location, shape and scale.
EXAMPLE 3 The unimodal continuous distributions.
NOTE 1 The class of distributions can often be fully specified through the values of the appropriate parameters.
2
ISO/TR 22514-4:2007(E)
NOTE 2 Tests for the normality of data can be found in ISO 5479.
NOTE 3 Adapted from ISO 3534-2:2006, 2.5.2.
2.1.9
distribution model
specified distribution (2.1.7) or class of distributions (2.1.8)
EXAMPLE 1 A model for the distribution of a product characteristic, such as the diameter of a bolt, might be the
normal distribution with mean 15 mm and standard deviation 0,05 mm. Here the model is a fully specified one.
EXAMPLE 2 A model for the diameter of bolts as in Example 1 could be the class of normal distributions without
attempting to specify a particular distribution. Here the model is the class of normal distributions.
[ISO 3534-2:2006, 2.5.3]
2.1.10
reference limits
nominated quantiles of the distribution (2.1.7) of the product characteristic
NOTE 1 The conditions of the distribution (2.1.7) of the product characteristic must be specified, see Notes 2 and 3 of
2.1.7.
NOTE 2 Traditionally the 0,135 % and 99,865 % quantiles have been used.
EXAMPLE If the distribution (2.1.7) of the product characteristic is normal with mean µ and standard deviation σ,
the limits are µ ± 3σ when the traditional 0,135 % and 99,865 % quantiles are used.
2.1.11
reference interval
interval bounded by the 99,865 % distribution quantile, X 99,865 % , and the 0,135 % distribution quantile,
X 0,135 %
NOTE 1 The interval can be expressed by ( X 99,865 %, X 0,135 % ) and the length of the interval is
( )
X 99,865 % − X 0,135 % .
NOTE 2 This term is used only as an arbitrary, but standardized, basis for defining the process performance index
(2.3.3) and process capability index (2.2.3).
NOTE 3 For a normal distribution (2.1.7), the reference interval may be expressed in terms of six standard deviations,
6σ, or 6S, when estimated from a sample.
NOTE 4 For a non-normal distribution, the length of the reference interval can be estimated by means of appropriate
probability papers (e.g. log-normal) or from the sample kurtosis and sample skewness using, for example, Pearson
curves.
NOTE 5 A quantile or fractile indicates division of a distribution into equal units or fractions, e.g. percentiles. Quantile is
defined in ISO 3534-1.
[ISO 3534-2:2006, 2.5.7]
2.1.12
upper fraction nonconforming
pU
fraction of the distribution (2.1.7) of a characteristic (2.1.2) that is greater than the upper specification limit
(2.1.4), U
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
EXAMPLE In a normal distribution (2.1.7), with mean, µ, and standard deviation σ :
⎛U − µ ⎞ ⎛ µ −U ⎞
pU = 1 − Φ ⎜ ⎟ =Φ ⎜ ⎟
⎝ σ ⎠ ⎝ σ ⎠
where
pU is the upper fraction nonconforming;
Φ is the distribution function of the standard normal distribution;
U is the upper specification limit.
NOTE 1 Tables (or functions in statistical computer packages) of the standard normal distribution are readily available
which give the proportion of process output expected beyond a particular value of interest, such as a specification limit
(2.1.4), in terms of standard deviations away from the process mean. This obviates the need to work out the statistical
distribution function given in the example.
NOTE 2 The function relates to a theoretical distribution. In practice, with empirical distributions, the parameters are
replaced by estimates.
[ISO 3534-2:2006, 2.5.4]
2.1.13
lower fraction nonconforming
pL
fraction of the distribution (2.1.7) of a characteristic (2.1.2) that is less than the lower specification limit
(2.1.4), L
EXAMPLE In a normal distribution (2.1.7), with mean, µ, and standard deviation, σ :
⎛L−µ ⎞
pL = Φ ⎜ ⎟
⎝ σ ⎠
where
pL is the lower fraction nonconforming;
L is the lower specification limit.
NOTE 1 Tables (or functions in statistical computer packages) of the standard normal distribution are readily available
which give the proportion of process output expected beyond a particular value of interest, such as a specification limit
(2.1.4), in terms of standard deviations away from the process mean. This obviates the need to work out the statistical
distribution function given in the example.
NOTE 2 The function relates to a theoretical distribution. In practice, with empirical distributions, the parameters are
replaced by estimates.
[ISO 3534-2:2006, 2.5.5]
4
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 22514-4:2007(E)
2.1.14
total fraction nonconforming
pt
sum of upper fraction nonconforming (2.1.12) and lower fraction nonconforming (2.1.13)
EXAMPLE In a normal distribution (2.1.7), with mean, µ, and standard deviation, σ :
⎛ µ −U ⎞ ⎛L−µ⎞
pt = Φ ⎜ ⎟ +Φ ⎜ ⎟
⎝ σ ⎠ ⎝ σ ⎠
where
pt is the total fraction nonconforming;
U is the upper specification limit;
L is the lower specification limit.
[ISO 3534-2:2006, 2.5.6]
2.2 Process capability, estimates and indices
2.2.1
process capability
〈estimate〉 statistical estimate of the outcome of a characteristic from a process which has been demonstrated
to be in a state of statistical control and which describes that process’s ability to realize a characteristic that
will fulfil the requirements for that characteristic
NOTE 1 The characteristic from the process needs to be documented to have been in statistical control.
NOTE 2 The outcome is a distribution (2.1.7), the class of which needs determination and its parameters estimated.
NOTE 3 In certain circumstances, the standard deviation, Sw, that represents only within subgroup variation can be
used as an estimator for St:
m m
∑ Sj ∑ Sj2
R j =1 j =1
Sw ≈ or or
d2 mc 4 m
where
R is the average range calculated from a set of m subgroup ranges;
Sj is the sample standard deviation of the jth subgroup;
m is the number of subgroups of the same size, n;

--`,,```,,,,````-`-`,,`,,`,`,,`---
d2, c4 are constants based on subgroup size, n.
Alternatively, for a normal distribution, the process overall standard deviation σt, can be estimated using the formula for St.
The value of the estimators St and Sw converge for a process in a state of statistical control. A comparison of the two gives
an indication of the degree of stability of the process. For an out-of-control process about a constant mean, or for a
process that is subject to systematic changes in the mean, the value of Sw is likely to significantly underestimate the
Copyright International Organization for Standardization © ISO 2007 – All rights reserved 5
ISO/TR 22514-4:2007(E)
process standard deviation. Hence Sw should be used with extreme caution. Sometimes, too, the estimator St is preferred
to Sw because it has more tractable statistical properties (e.g. facilitating the calculation of confidence limits).
NOTE 4 For a normal distribution, process capability can be assessed from the expression:
X ± zS t
where
m
1
X =
m ∑X j
j =1
and X j is the observed mean of the jth subgroup.
The choice of z depends on the particular parts per million capability standard used. Typically z takes the value of 3, 4 or 5.
If the process capability meets the specified requirements, a z value of 3 indicates an expected 2 700 parts per million
outside of specification. Similarly, a z of 4 indicates an expected 64 parts per million and a z of 5 an expected 0,6 parts per
million outside of specification.
NOTE 5 For a non-normal distribution, process capability can be assessed using, for example, an appropriate
probability paper or from the parameters of the distribution fitted to the data. The expression for process capability takes
the asymmetric form:
X −+ba
The notation +−ba is in the same style as standard drawing office practice for expressing specified tolerances about a
nominal, or preferred, value for a characteristic when the preferred value is not equidistant from each limit. The equivalent
notation for limits symmetrical about the preferred value is ±. This enables a direct comparison to be made between the
dimensional performance of a characteristic and its specified requirements in terms of both location and dispersion.
R
NOTE 6 When S w = is used, it needs to be appreciated that this estimator:
d2
⎯ becomes progressively less efficient as the subgroup size increases;
⎯ is very sensitive to the distribution of individual values;
⎯ makes it difficult to estimate confidence limits.
NOTE 7 Capability conditions are very restrictive and include:
⎯ the methods applied to demonstrate that the process is in control;
⎯ technical conditions (input batches, operators, tools, etc.);
⎯ the measurement process (discrimination, accuracy, repeatability, reproducibility, etc.); and
⎯ data collection (duration, frequency).
2.2.2
process capability estimate
quantity that describes one or more properties of the distribution (2.1.7) of the product characteristic under
capability conditions
EXAMPLE 1 The standard deviation (ISO 3534-1:2006, 2.37) of the distribution (2.1.7) of the product characteristic
under capability conditions (see 2.2.1, Notes 1 and 7).
--`,,```,,,,````-`-`,,`,,`,`,,`---
6
ISO/TR 22514-4:2007(E)
EXAMPLE 2 The mean (ISO 3534-1:2006, 2.35.1) of the distribution (2.1.7) of the product characteristic under
capability conditions (see 2.2.1, Notes 1 and 7).
EXAMPLE 3 The reference interval (2.1.11) of the distribution (2.1.7) of the product characteristic under capability
conditions (see 2.2.1, Notes 1 and 7).
2.2.3
process capability index
Cp
index describing process capability (2.2.1) in relation to specified tolerance (2.1.5)
NOTE 1 Frequently, the process capability index is expressed as the value of the specified tolerance (2.1.5) divided
by a measure of the length of the reference interval (2.1.11) for a process in a state of statistical control, namely as:
U −L
Cp =
X 99,865 % − X 0,135 %
NOTE 2 For a normal distribution (2.1.7), the reference interval (2.1.11) is equal to 6Sw (see 2.2.1, Notes).
NOTE 3 For a non-normal distribution (2.1.7), the reference interval (2.1.11) can be estimated, for example, using
the probability paper method or the Pearson curves method.
2.2.4
upper process capability index
CpkU
index describing process capability (2.2.1) in relation to the upper specification limit (2.1.4), U
NOTE 1 Frequently, the upper process capability index is expressed as the difference between the upper
specification limit (2.1.4) and the 50 % distribution quantile, X 50 % divided by a measure of the length of the upper
reference interval (2.1.11) for a process in a state of statistical control, namely as:
U − X 50 %
C pkU =
X 99,865 % − X 50 %
NOTE 2 For a normal distribution (2.1.7), the upper reference interval (2.1.11) is equal to 3Sw (see 2.2.1, Notes) and
X 50 % represents both the mean and the median.
NOTE 3 For a non-normal distribution (2.1.7), the upper reference interval (2.1.11) can be estimated, for example,
using the probability paper method or the Pearson curves method and X 50 % represents the median.
2.2.5
lower process capability index
CpkL
index describing process capability (2.2.1) in relation to the lower specification limit (2.1.4)
NOTE 1 Frequently, the lower process capability index is expressed as the difference between the 50 % distribution
quantile, X 50 % , and the lower specification limit (2.1.4) divided by a measure of the length of the lower reference
interval (2.1.11) for a process in a state of statistical control, namely as:
X 50 % − L
C pkL =
X 50 % − X 0,135 %
NOTE 2 For a normal distribution (2.1.7) the lower reference interval (2.1.11) is equal to 3Sw (see 2.2.1, Notes) and
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
NOTE 3 For a non-normal distribution (2.1.7), the lower reference interval (2.1.11) can be estimated, for example,
2.2.6
minimum process capability index
Cpk
smaller of upper process capability index (2.2.4) and lower process capability index (2.2.5)
[ISO 3534-2:2006, 2.7.5]

--`,,```,,,,````-`-`,,`,,`,`,,`---
2.3 Process performance, measures and indices
2.3.1
process performance
〈measure〉 statistical measure of the outcome of a characteristic from a process that may not have been
demonstrated to be in a state of statistical control
NOTE 1 The characteristic from the process does not need to be documented to have been in statistical control.
NOTE 2 The outcome is a distribution (2.1.7) the class of which needs determination and its parameters assessed.
NOTE 3 Care should be exercised in using this measure as it may contain a component of variability due to special
causes the value of which is not predictable.
NOTE 4 For a normal distribution (2.1.7) described in terms of the standard deviation, St, estimated from only one
sample of size N, the standard deviation is expressed thus:
N
∑( X i − X )
2
St = i =1
N −1
where
N
∑ Xi
X = i =1
N
The descriptor, St, takes into account the variation both due to random (common) causes together with any special causes
that may be present. St is used here instead of σt as the standard deviation is a statistical descriptive measure. The
sample size can be made up of m subgroups each of size n.
NOTE 5 For a normal distribution (2.1.7) process performance can be assessed from the expression:
X ± zS t
The choice of z depends on the particular parts per million performance requirement. Typically z takes the value of 3, 4 or
5. If the process capability meets the specified requirements, a z value of 3 indicates an expected 2 700 parts per million
outside of specification. Similarly, a z of 4 indicates an expected 64 parts per million and a z of 5 an expected 0,6 parts per
million outside of specification.
NOTE 6 For a non-normal distribution, process performance can be assessed using, for example, an appropriate
probability paper or from the parameters of the distribution fitted to the data. The expression for process performance
takes the asymmetric form:
X −+ba
8
ISO/TR 22514-4:2007(E)
The notation +−ba is in the same style as standard drawing office practice for expressing specified tolerances about a
nominal, or preferred, value for a characteristic when the preferred value is not equidistant from each limit. The equivalent
notation for limits symmetrical about the preferred value is ±. This enables a direct comparison to be made between the
dimensional performance of a characteristic and its specified requirements in terms of both location and dispersion.
NOTE 7 Performance conditions are least restrictive, but include:
⎯ technical conditions (input batches, operators, tools, etc.);
⎯ the measurement process (discrimination, accuracy, repeatability, reproducibility, etc.); and
⎯ data collection (duration, frequency).
2.3.2
process performance measure
quantity that describes one or more properties of the distribution (2.1.7) of the product characteristic under
performance conditions
EXAMPLE 1 The standard deviation (ISO 3534-1:2006, 2.37) of the distribution (2.1.7) of the product characteristic
under performance conditions (see 2.3.1, Notes 1 and 7).
EXAMPLE 2 The mean (ISO 3534-1:2006, 2.35.1) of the distribution (2.1.7) of the product characteristic under
performance conditions (see 2.3.1, Notes 1 and 7).
EXAMPLE 3 The reference interval (2.1.11) of the distribution (2.1.7) of the product characteristic under
--`,,```,,,,````-`-`,,`,,`,`,,`---
performance conditions (see 2.3.1, Notes 1 and 7).
2.3.3
process performance index
Pp
index describing process performance (2.3.1) in relation to specified tolerance (2.1.5)
NOTE 1 Frequently, the process performance index is expressed as the value of the specified tolerance (2.1.5)
divided by a measure of the length of the reference interval (2.1.11) for a process in a state of statistical control, namely
as:
U −L
Pp =
X 99,865 % − X 0,135 %
NOTE 2 For a normal distribution (2.1.7), the length of the reference interval (2.1.11) is equal to 6St (see 2.3.1,
Notes).
NOTE 3 For a non-normal distribution (2.1.7), the length of the reference interval (2.1.11) can be estimated, for
example, using the probability paper method or the Pearson curves method.
2.3.4
upper process performance index
PpkU
index describing process performance (2.3.1) in relation to the upper specification limit (2.1.4), U
NOTE 1 Frequently, the upper process performance index is expressed as the difference between the upper
specification limit (2.1.4) and the 50 % distribution quantile, X 50 %, divided by a measure of the length of the upper
reference interval (2.1.11), namely as:
U − X 50 %
PpkU =
X 99,865 % − X 50 %
ISO/TR 22514-4:2007(E)
NOTE 2 For a normal distribution (2.1.7), the upper reference interval (2.1.11) is equal to 3St (see 2.3.1, Notes) and
NOTE 3 For a non-normal distribution (2.1.7), the upper reference interval (2.1.11) can be estimated, for example,
2.3.5
lower process performance index
PpkL
index describing process performance (2.3.1) in relation to the lower specification limit (2.1.4), L
NOTE 1 Frequently, the lower process performance index is expressed as the difference between the 50 % distribution
quantile, X 50 %, and the lower specification limit (2.1.4) divided by a measure of the length of the lower reference
interval (2.1.11), namely as:
X 50 % − L
PpkL =
X 50 % − X 0,135 %
NOTE 2 For a normal distribution (2.1.7), the lower reference interval (2.1.11) is equal to 3St (see 2.3.1, Notes) and
NOTE 3 For a non-normal distribution (2.1.7), the lower reference interval (2.1.11) can be estimated, for example,
2.3.6
minimum process performance index
Ppk
smaller of upper process performance index (2.3.4) and lower process performance index (2.3.5)
--`,,```,,,,````-`-`,,`,,`,`,,`---
[ISO 3534-2:2006, 2.6.5]
3 Symbols and abbreviated terms
3.1 Symbols
In addition to the symbols listed below, some symbols are defined where they are used within the text.
α fraction or proportion
β shape parameter in a Weibull distribution
β2 coefficient of kurtosis
c number of nonconformities in a sample of size n
c average number of nonconformities in a sample of size n
c4 constant based on subgroup size, n (see ISO 8258)
Cp process capability index
Cpk minimum process capability index
CpkL lower process capability index
10
ISO/TR 22514-4:2007(E)
CpkU upper process capability index
CR process capability fraction (PCF)
d2 constant based on subgroup size, n (see ISO 8258)
e mathematical constant
Φ distribution function of the standard normal distribution
γ location parameter in a Weibull distribution
γ1 coefficient of skewness
m number of subgroups
Kl, Ku multipliers for estimating the confidence limits for a process capability index
L lower specification limit
P0,135 % lower 0,135 % quantile for the Pearson system of curves
µ location of the process; population mean value
N total sample size
n number of values or subgroup size (for a control chart)
np number nonconforming in a sample of size n

--`,,```,,,,````-`-`,,`,,`,`,,`---
np average number nonconforming in a sample of size n
Pα % α percentile
p proportion of nonconforming items in a sample
p average proportion of nonconforming items in a sample
pL lower fraction nonconforming
Pp process performance index
Ppk minimum process performance index
PpkL lower process performance index
PpkU upper process performance index
pt total fraction nonconforming
pU upper fraction nonconforming
P99,865 % upper 99,865 % quantile for the Pearson system of curves
π geometric constant

ISO/TR 22514-4:2007(E)
Qk process variation index
θ parameter required for the Rayleigh distribution
R average subgroup range
S standard deviation, sample statistic
St standard deviation, with the subscript ‘t’ indicating total
S average sample standard deviation
Sj observed sample standard deviation of the jth subgroup
σ standard deviation, population
σˆ t estimated standard deviation, total
T target value
U upper specification limit
u number of nonconformities per item in a subgroup
u average number of nonconformities per item in a subgroup
Xα % α % distribution quantile
Xi ith value in a sample
X arithmetic mean value, sample
X arithmetic mean, of a number of sample arithmetic means
ξ scale parameter in a Weibull distribution
Y1, Y2 values read from a graph
zα quantile of the standardized normal distribution from −∞ to α
3.2 Abbreviated terms
FRC first run capability
MSE mean square error
NHU nonconformities per hundred units
NMU nonconformities per million units
PCF process capability fraction
PCI process capability indices
--`,,```,,,,````-`-`,,`,,`,`,,`---
12
ISO/TR 22514-4:2007(E)
4 Statistical measures used in process capability and performance
4.1 General
The measures referred to below in 4.2 to 4.6 refer only to measured data. They are unsuitable for count or
attribute data and information concerning the expression of measures for such data will be found later in 4.7.
4.2 Measures of location
The most common measure of location used is the mean, µ, although the sample median, X 50 % , is
sometimes used. With non-normally distributed process measurements, the median is often a preferred
alternative measure.
4.3 Measures of dispersion

--`,,```,,,,````-`-`,,`,,`,`,,`---
4.3.1 Inherent variability
The preferred measure to quantify inherent process variation (ISO 3534-2:2006, 2.2.2) is the standard
deviation, σ. This is often estimated from the mean range value, R , taken from a range (R) chart when the
process is stable and in a state of statistical control as indicated in 5.1. Methods used to estimate the process
standard deviation are given in Annex A.
4.3.2 Total variability
It is necessary to differentiate between a standard deviation that measures only short-term variation and that
which measures longer-term variation. Methods of calculating the standard deviations representing these
variations are given in Annex A.
Very often, when data are gathered over a long period of time, the standard deviation is made larger by the
effects of fluctuations in the process. When this standard deviation is computed, the symbol σˆ t is used within
this part of ISO 22514.
4.4 Mean square error (MSE)
When minimizing variation some practitioners, use the mean square error as the preferred measure. It is
compatible with the methods used in off-line quality techniques.
4.5 Reference limits
The lower and upper reference limits are respectively defined as the 0,135 % and the 99,865 % quantiles of
the distribution that describe the output of the process characteristic. They are written as X 0,135 % and
X 99,865 % .
4.6 Reference interval, also known as process spread
The reference interval is the interval between the upper and the lower reference limits. The reference interval
includes 99,73 % of the individuals in the population from a process that is in a state of statistical control.
4.7 Attributes
Measurement of quality by attributes consists of recording the presence (or absence) of some characteristic or
attribute in each of the items in the subgroup under consideration. Counts are made of how many items do (or
do not) possess the quality attribute or how many such events occur in the item, group of items or unit area.
ISO/TR 22514-4:2007(E)
When the data are attributive, the usual statistic will be the number of nonconforming items (np) or the number
of nonconformities (c) found. Sometimes, it will be necessary to compute the fraction nonconforming (p) or the
number of nonconformities per item (u) depending upon the sampling strategies adopted.
5 Capability
5.1 General
Process capability is a measure of inherent process variability. The variability that is inherent in a process
when operating in a state of statistical control is known as the inherent process variability. It represents the
variation that remains after all known removable assignable causes have been eliminated. If the process is
--`,,```,,,,````-`-`,,`,,`,`,,`---
monitored using a control chart, the control chart will show an in control state.
Capability is often regarded as being related to the proportion of output that will occur within the product
specification tolerances. Since a process in statistical control can be described by a predictable distribution,
the proportion of out-of-specification outputs can be estimated. As long as the process remains in statistical
control, it will continue to produce the same proportion out-of-specification.
Management actions to reduce the variation from random causes are required to improve the process' ability
to consistently meet the specification requirements.
In short, the following will be necessary:
a) define the process and its operating conditions. If there is a change to those conditions, it will necessitate
a new process study;
b) assess the short-term and long-term measurement variabilities as percentages of the total variability and
minimize them;
c) preserve the process stability and maintain its statistical control;
d) estimate the remaining inherent variation; and
e) select an appropriate measure of capability.
The following are the conditions that will apply for capability:
⎯ all technical conditions, e.g. temperature and humidity, must be clearly stated;
⎯ the uncertainty of the measurement system must be specified;
⎯ multi-factor, multi-level aspects of the process should be allowed;
⎯ the duration over which the data has been gathered must be recorded;
⎯ the frequency of sampling must be specified and the start and finish dates of data collection;
⎯ the process must be controlled with a control chart; and
⎯ the process must be in a state of statistical control.
It is necessary to check the control chart from which the data have been taken for statistical control and to
examine a histogram of the data with any specification limits superimposed upon it. A valid test for normality
should be used in assessing the data such as the Anderson-Darling test [8]. This test is powerful in detecting
departures from normality in the tails of the distribution and is suggested here as this is the region of interest
14
ISO/TR 22514-4:2007(E)
for capability and performance indices. Additionally, normal probability paper can be used to look for the
following:
1) verification of normality;
2) outliers;
3) data beyond any specification limit;
4) whether the data are well inside the specification limit(s);
5) evidence of asymmetry (i.e. skewness);
6) evidence of “long tails” in the data (i.e. kurtosis);
7) off-centre distribution; and
8) any unusual features.
Explanations of anomalies should be sought in relation to these mentioned features and appropriate action
taken on the data prior to the calculation of any measure. It would be inappropriate to just discard data that do
not appear to fit any preconceived pattern. Such departures might be very revealing about the process's
behaviour and should be thoroughly investigated.
5.2 Process capability
5.2.1 Normal distribution
Process capability is defined as a statistical measure of inherent process variability for a given characteristic.
The conventional method is to take the reference interval which describes 99,73 % of the individual values
from a process that is in a state of statistical control with the 0,135 % remaining on each side. This applies
even if the population of individual values is not normally distributed. For a normal distribution, this process
--`,,```,,,,````-`-`,,`,,`,`,,`---
interval is represented by six standard deviations. See Figure 1.
Key
a Reference interval 99,73%.
Figure 1
On occasions, process capability is taken to account for extra sources of variation such as a multiple stream
process, for example, output from a multi-cavity injection moulding press. Under these circumstances, the
distribution of values may still be approximately normal but with extra variability so that the standard deviation
must represent the total variation, σ t . It is important to state how the standard deviation has been calculated
as well as the sampling strategy used, sample size and the quantity and variability of output produced
between samples as these will in practice affect the validity of the capability assessment.

ISO/TR 22514-4:2007(E)
Data will usually be taken from a control chart. If the control chart had relaxed control lines or modified control
lines, the real process standard deviation will be larger than that estimated from data taken from a control
chart with standard control lines. Issues such as these and those given earlier will influence the reference
interval and it is important that they are stated in any capability assessment.
“Capable” processes will be those whose reference intervals are less than any specified tolerance by a
particular amount. An example of this is shown in Figure 2.
Key
a Reference interval 99,73 %.
Figure 2
--`,,```,,,,````-`-`,,`,,`,`,,`---
5.2.2 Non-normal distribution
If the distribution of individual values does not form a normal distribution but is skewed, then the reference
interval may appear as in Figure 3. The values Y1 and Y2, which will usually be the 0,135 % and the 99,865 %
quantiles, can be estimated using a suitable probability paper (see Figure 4 for an example using an extreme
value distribution probability paper) or by the use of suitable computer software (see Annex E). They may also
be computed using tabular values (see Annex B) or using the particular probability function as suggested in
Annex C.
Key
a Reference interval 99,73 %.
Figure 3
5.3 Process location
Even if a process can be deemed capable by the above definition (in 5.2.1), if the process distribution has
been poorly centred relative to the specification limits, out-of-specification items may be produced. For this
reason, it is necessary to assess the location in addition to the process interval.
16
ISO/TR 22514-4:2007(E)
Key
best fit line
cumulative percent
Figure 4
5.4 Process capability indices for measured data
5.4.1 General
It should be noted that when the capability indices given in this part of ISO 22514 are computed, they only
form point estimates of their true values. It is therefore recommended that wherever possible the indices'
confidence intervals are computed and reported. Methods by which these can be computed are described in
Annex D.
It is effective to express the capability of a process with the use of an index number. Several indices are given.
Care must be taken when handling non-normal distributions.
The process capability index often used is the ratio of a specified tolerance to the reference interval and is
known as Cp. Thus:
U −L
Cp = (1)
X 99,865% − X 0,135%
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
where
L is the lower specification limit;
U is the upper specification limit.
There are other indices that incorporate both the location and the variation; of these, the most widely used is
the Cpk index. If the index is less than a given value, the process is deemed to be producing an excessive
proportion of items outside of the specification.
The Cpk index is the ratio of the difference between a specified tolerance limit and the process location to the
difference between the corresponding natural process limit and the process location.
U − X 50 %
CpkU = (2)
X 99,865% − X 50%
and
X 50% − L
CpkL =
X 50% − X 0,135%
where
L is the lower specification limit;
U is the upper specification limit;
X 50% is the median value.
The Cpk index is reported as the smaller value of these.
Some practitioners report both of the above values that are also known as CPU and CPL, respectively. This
provides information about both sides of the process.
These indices will provide information about whether a process is poorly centred and whether it will possibly
produce out-of-specification items. Even if the Cp index is high, a low value of the Cpk index will reveal a poorly
centred process and a high probability of producing out-of-specification items.
5.4.2 The Cp index (normal distribution)
If the individual values form a normal distribution, the length of the reference interval is equal to 6σ, where σ is
the inherent process standard deviation. Therefore, the Cp index can be expressed as:
U −L
Cp =
6σ
An estimate ( σˆ ) of the inherent process standard deviation (σ ) is required to obtain an estimate of the Cp
index. When this has been obtained, usually with data from a control chart once the process is shown to be
statistically stable (see 5.1), the index is estimated:
U −L
Cˆ p =
6σˆ
--`,,```,,,,````-`-`,,`,,`,`,,`---
18
ISO/TR 22514-4:2007(E)
5.4.3 The Cpk index (normal distribution)
When the distribution of individual values forms a normal distribution, the median X 50 % is equal to the mean
(µ). Further, X 99,865 % − X 50 % and X 50 % − X 0,135 % are each equal to 3σ. Therefore, the Cpk index can be
expressed as the minimum of:
U −µ
C pkU =
3σ
or
µ−L
C pkL =
3σ
Cpk may be estimated by the minimum of:
U−X
Cˆ pkU =
3σˆ
or
X −L
Cˆ pkL =
3σˆ
In computing a capability index, thought must be given to the measure of the process variation used in the
denominator. Here, σ is given to represent the variation when the data comes from a process that is in a state
of statistical control.
The data might come from a multiple stream process such as a multi-headed filling machine or a multi-spindle
machine where the total output is treated together, where data from all streams are simultaneously
considered. The lower the index, the greater the proportion of items produced out-of-specification.
5.5 Process capability indices for measured data (non-normal)
5.5.1 General
If the distribution of individual values is non-normal, the expressions in Equations (1) and (2) still apply but the
estimation of the indices becomes more complicated. Three approaches to estimate the reference limits are
--`,,```,,,,````-`-`,,`,,`,`,,`---
given.
The probability paper method described in 5.5.2 is fairly simple and requires little computation but is
somewhat crude. The approach given in 5.5.4 is computationally more involved but is superior to any other
method as far as accuracy is concerned.
5.5.2 Probability paper method
From graphs similar to that shown in Figure 4, estimates of the quantiles X 0,135 % and X 99,865 % can be
obtained. The estimates are denoted by Y1 and Y2 respectively and Equation (1) becomes:
U −L
Cˆ p =
Y2 − Y1

ISO/TR 22514-4:2007(E)
In a similar way, the Cpk formulae become:
U − X 50 %
Cˆ pkU =
Y2 − X 50 %
or
X 50 % −L
Cˆ pkL =
--`,,```,,,,````-`-`,,`,,`,`,,`---
X 50 % − Y1
whichever gives the lower value. If the index is less than a given value, the process is deemed to be
producing an excessive proportion of items outside of the specification. The proportion nonconforming
depends upon the distribution and the value of the index. The link between the index and the proportion of
nonconforming items produced depends on the class of distributions. Care should be taken not to interpret
indices on the basis of cut-off points that have been derived for the normal distribution and hence are only
applicable for that distribution.
Note that the probability paper method directly estimates fairly extreme quantiles and this can be inaccurate.
5.5.3 Pearson curves method
As an alternative to using probability paper, standardized Pearson curves may be used. The method is
described by way of an example (see Annex B). The index is computed using the formula:
U −L
Cˆ p =
Xˆ 99,865 % − Xˆ 0,135 %
where X̂ 0,135 % and X̂ 99,865 % are the 0,135 % and 99,865 % quantiles estimated from the standardized
Pearson curves.
Also, we have the formulae:
U − Xˆ 50 %
Cˆ pkU =
Xˆ 99,865 % − Xˆ 50 %
or
Xˆ 50 % − L
Cˆ pkL =
Xˆ 50 % − Xˆ 0,135 %
where X̂ 50 % is the estimated median.
In order to use this method, it is necessary to establish skewness and kurtosis values in addition to the mean
and standard deviation for the data set upon which the index is to be computed.
This method is not preferred but is presented here for completeness due to its occasional use.
This approach, and a similar one based on Johnson curves, should be regarded with considerable caution,
especially when it is a procedure within a “black box” computer program used to analyse large sets of data.
Some of the potential difficulties are as follows:
⎯ within a system of distributions, some distributions are more difficult to fit than others. The method of
moments can yield unstable or inefficient curve parameters in some cases;
20
ISO/TR 22514-4:2007(E)
⎯ unless the estimation technique is applied skilfully, it is possible to obtain fitted curves that are
meaningless over certain ranges of the data. For example, with the method of moments, an easily made
mistake is to fit a Pearson Type III distribution whose estimated threshold is less than the lower bound for
the process output, thereby invalidating the estimates of X 0,135 % and Cpk;
⎯ the method of moments does not yield estimates of the variability in the estimated indices. Likewise,
these methods do not yield confidence intervals for the indices;
⎯ not every data distribution can be described adequately with a Pearson or Johnson curve;
⎯ goodness-of-fit tests are limited to the chi-squared test since more powerful tests are not generally
available for the Pearson and Johnson systems;
⎯ the “black box” approach tends to displace basic practices, such as plotting the data and applying simple
normalizing transformations, that provide genuine understanding of the process.
5.5.4 Distribution identification method
Annex C describes certain families of distribution functions (such as the log-normal distribution, the Rayleigh
and the Weibull distributions) that are commonly found when investigating process capability. The method is
first to identify the appropriate family of distributions, secondly to estimate the parameters of the distribution of
the family that best explain the data by some efficient estimation method and finally to express the quantiles in
terms of the parameters of that distribution.
This is analogous to the procedure adopted in the case of the normal distribution where σ is estimated and 6σ
is represented by ( X 99,865 % − X 0,135 % ) .
Various types of probability paper may be useful to identify the appropriate family of distributions.
5.6 Alternative method for describing and calculating process capability estimates
The bases for this method are the widely used definitions of Cp and Cpk for the “ideal process” with a normally
distributed characteristic X, where the expectation µ and variance σ 2 are constant with time and the
corresponding estimates are X and S 2 .
Table 1 — Process capability indices and estimates (normal distribution)
Index Estimate
U −L U −L
Cp = Cˆ p =
6σ 6S
U −µ U−X
C pkU = Cˆ pkU =
3σ 3S
µ−L
C pkL = X −L
3σ Cˆ pkL =
3S
Cpk = min(CpkL , C pkU )
C = min(Cˆ
ˆ
pk
ˆ
pkL , C pkU )
This “ideal process” implies that the long-term standard deviation is equal to the short-term standard deviation.
When these measures of process capability have to be extended to characteristics that are not normally
distributed, it is necessary to begin from the point of view that these measures are meant to be managerial
tools, that they display and reflect the conformance of the actual values of a characteristic with its
corresponding specification limits. Therefore, these measures must be linked to the fraction of actual values
conforming or nonconforming. In particular, the same fraction conforming or nonconforming has to yield the
21
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
same values of capability or performance independent of the shape of the distribution of the actual values of a
characteristic.
The following formulae are equivalent to those in Table 1:
Table 2 — Process capability indices and estimates (normal distribution) — Equivalent formulae
Index Estimate
C pkU + CpkL Cˆ pkU + Cˆ pkL

Cp = Cˆ p =
2 2
z1− pU z1− pˆ U
C pkU = Cˆ pkU =
3 3
z1− p L z1− pˆ L
C pkL = Cˆ pkL =
3 3
where pU and p L are the fractions nonconforming at the upper and lower specification limits and pˆ U , pˆ L
are the corresponding estimates. The formulae in the above table can be applied to any distribution.
It is assumed that the user has knowledge of the shape of the distribution because of what is known about the
manufacturing process or by some evaluation of a sample by an appropriate probability paper.
For those distributions that are frequently observed (normal, log-normal, Rayleigh and Weibull), the required
relations and formulae are given in Annex C.
5.7 Other capability indices for measured data in other special cases
5.7.1 Process capability fraction (PCF)
The PCF is the inverse of the Cp index:
6σ 1
=
U − L Cp
It can be expressed as a percentage value and occasionally named CR (%).
5.7.2 Indices when the specification limit is one-sided or no specification limit is given
5.7.2.1 General
Sometimes, a specification is given that has only one limit, e.g. a maximum value. In these circumstances, it
will only be possible to compute a Cpk or a Ppk index.
There will also be situations when specification limits are not given or not known. However, if a target or
nominal value is given for the product characteristic or process parameter the following measures might be
appropriate. They present a special appeal to those engaged in minimizing process variation around a target
value.
5.7.2.2 Mean square error (MSE)
The mean square error provides a measure that involves both location and variation. It is computed as
follows:
σ 2 + (µ − T )
2
--`,,```,,,,````-`-`,,`,,`,`,,`-
22
ISO/TR 22514-4:2007(E)
where
σ is the standard deviation of the process;
µ is the achieved mean value of the process;
T is the given target value for the process.
In deriving this measure from data, it is necessary to provide estimates of the process standard deviation and
µ using sample data from a control chart.
5.7.2.3 Qk index
This index uses the mean square error given in 5.7.2.2, but expresses the whole value as a coefficient of
variation and is computed as follows:
100 σ 2 + ( µ − T )
2
Qk = (%)
T
for T ≠ 0 .
An interesting property of this index is if the process drifts from its target, the index will increase in value, and
if the process variation increases, it will also increase the value of the index. The smaller this index becomes,
the better the process is deemed to have performed.
5.8 Assessment of proportion out-of-specification (normal distribution)
The proportion of out-of-specification items ( p L and pU ) that fall below L and above U can be estimated
using properties of the standard normal distribution. Standardized deviates can be calculated as follows:
z pˆ = 3Cˆ pkU
U
and
z pˆ L = 3Cˆ pkL
pˆ U and pˆ L are found as the proportions exceeding z p and z p L , respectively, in a standard normal
U
distribution.
Additionally, the process yield can be computed as 100 % minus the total percentage nonconforming, in the
case of a controlled process.
If a characteristic, in statistical control and stable, has a CpkU of 0,86 and a CpkL of 0,91, the proportion of
out-of-specification can be determined using the method given above as follows.
a) Calculate the “upper” standardized deviate:
z pˆ L = 3Cˆ pkL
= 3 × 0,91
= 2,73
b) Calculate the “lower” standardized deviate:
z pˆ L = 3Cˆ pkU
= 3 × 0,86
= 2,58
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
c) Using a standard normal table, look up the values pˆ U and pˆ L for the proportions of the distribution
beyond the specification limits U and L, z p and z p L respectively.
U
For convenience and ease of use, Table 3 gives look-up values for the estimated proportion out-of-
specification. Table 3 is indexed by CpkU or CpkL , the process capability indices (PCI). Table 3 should not be
used to derive Cp nor Cpk values for attribute data.
Using the above example of a CpkU of 0,86 and a CpkL of 0,91, the estimated proportion beyond the
specification limits U and L may be read directly from Table 3 as 0,004 9 and 0,003 2.
Table 3 — CpkU or CpkL (PCI in the table) and proportion of the normal distribution remaining
in the tails of the distribution beyond a specification limit
PCI 0,00 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08 0,09
1,6 7,9 × 10−07 6,8 × 10−07 5,9 × 10−07 5,0 × 10−07 4,3 × 10−07 3,7 × 10−07 3,2 × 10−07 2,7 × 10−07 2,3 × 10−07 2,0 × 10−07
1,5 3,4 × 10−06 3,0 × 10−06 2,6 × 10−06 2,2 × 10−06 1,9 × 10−06 1,7 × 10−06 1,4 × 10−06 1,2 × 10−06 1,1 × 10−06 9,2 × 10−07
1,4 1,3 × 10−05 1,2 × 10−05 1,0 × 10−05 8,9 × 10−06 7,8 × 10−06 6,8 × 10−06 5,9 × 10−06 5,2 × 10−06 4,5 × 10−06 3,9 × 10−06
1,3 4,8 × 10−05 4,2 × 10−05 3,7 × 10−05 3,3 × 10−05 2,9 × 10−05 2,6 × 10−05 2,3 × 10−05 2,0 × 10−05 1,7 × 10−05 1,5 × 10−05
1,2 0,000 2 0,000 1 0,000 1 0,000 1 0,000 1 0,000 1 0,000 1 0,000 1 0,000 1 0,000 1
1,1 0,000 5 0,000 4 0,000 4 0,000 3 0,000 3 0,000 3 0,000 3 0,000 2 0,000 2 0,000 2
1,0 0,001 3 0,001 2 0,001 1 0,001 0 0,000 9 0,000 8 0,000 7 0,000 7 0,000 6 0,000 5
0,9 0,003 5 0,003 2 0,002 9 0,002 6 0,002 4 0,002 2 0,002 0 0,001 8 0,001 6 0,001 5
0,8 0,008 2 0,007 5 0,006 9 0,006 4 0,005 9 0,005 4 0,004 9 0,004 5 0,004 1 0,003 8
0,7 0,017 9 0,016 6 0,015 4 0,014 3 0,013 2 0,012 2 0,011 3 0,010 4 0,009 6 0,008 9
0,6 0,035 9 0,033 6 0,031 4 0,029 4 0,027 4 0,025 6 0,023 9 0,022 2 0,020 7 0,019 2
0,5 0,066 8 0,063 0 0,059 4 0,055 9 0,052 6 0,049 5 0,046 5 0,043 6 0,040 9 0,038 4
0,4 0,115 1 0,109 3 0,103 8 0,098 5 0,093 4 0,088 5 0,083 8 0,079 3 0,074 9 0,070 8
0,3 0,184 1 0,176 2 0,168 5 0,161 1 0,153 9 0,146 9 0,140 1 0,133 5 0,127 1 0,121 0
0,2 0,274 3 0,264 3 0,254 6 0,245 1 0,235 8 0,226 6 0,217 7 0,209 0 0,200 5 0,192 2
0,1 0,382 1 0,370 7 0,359 4 0,348 3 0,337 2 0,326 4 0,315 6 0,305 0 0,294 6 0,284 3
0,0 0,500 0 0,488 0 0,476 1 0,464 1 0,452 2 0,440 4 0,428 6 0,416 8 0,405 2 0,393 6
5.9 Attributes
5.9.1 General
It is often the case with attribute processes that the objective is zero nonconformities or zero nonconforming
items. Consequently, process capability in the case of a controlled process is restricted to statements about
the level of nonconformities or nonconforming items. This usually takes the form of the process average, for
example, p, and the indices given above and Table 3 do not apply.
If attributes data are generated by a measurement system that evaluates a variables measurement as a pass
or fail judgement, care must be taken to minimize any measurement system variation.
5.9.2 Capability measures and indices in the case of nonconforming units (np) or proportion of
nonconforming units (p)
When a process is monitored using a np-chart or a p-chart (see ISO 7870-1 for a description), the process
capability can be described by the average level p or np once the process is shown to be statistically stable.
--`,,```,,,,````-`-`,,`,,`,`,,`---
24
ISO/TR 22514-4:2007(E)
The process yield, sometimes called the first run capability (FRC), can be computed as shown in Equation (3),
i.e. FRC is the percentage of satisfactory items produced:
100 (1 − p ) (%)
or
⎛ np ⎞
100 ⎜ 1 − (%) (3)
⎝ n ⎟⎠
5.9.3 Capability measures and indices in the case of number of nonconformities (c) or proportion of
nonconformities (u)
When a process is monitored using a c-chart or a u-chart (see ISO 7870-1 for a description), the process
performance should be expressed by recording the average level, c or u , once the process is shown to be
statistically stable.
Additionally, the rate of occurrence of nonconformities can be computed as the nonconformities per hundred
units (NHU):
⎛c ⎞
100 ⎜ ⎟
⎝n⎠
or
100u
where n is the subgroup size taken.
If the occurrence of nonconformities per hundred is so low that the NHU is very much less than the value 1,
consider the number of nonconformities per million instead, i.e. NMU. When processing discrete items, this
measure is often given as parts per million (ppm).
The NHU is a useful measure when comparing processes under different subgroup sizes and provides
information concerning the expected rate of the nonconformity in the output from the process.
6 Performance
6.1 General
Process performance for a characteristic is the achieved distribution of results. The single important difference
between performance and capability is that for performance there is no requirement for the process to be in a
state of statistical control nor for the process to be controlled using a control chart. The following are the
conditions that will apply for performance:
⎯ all technical conditions, e.g. temperature and humidity, must be clearly stated;
⎯ the uncertainty of the measurement system must be specified;
⎯ multi-factor, multi-level aspects of the process should be allowed;
⎯ the duration over which the data has been gathered must be recorded;
⎯ the frequency of sampling must be specified and the start and finish dates of data collection;
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
⎯ the process need not be controlled with a control chart;
⎯ the process need not be in a state of statistical control, in particular historical data where the sequence is
unknown can be used to evaluate process performance.
Indices are given below to express process performance. Their form is similar to those already given in the
clause on capability and the general relationships given in Equations (1) and (2) for measured data are used
except they are named Pp , PpkU and PpkL respectively.
6.2 Process performance indices for measured data (normal distribution)
6.2.1 The Pp index
When the individual values form a normal distribution, the length of the reference interval is equal to 6σ t
where σ t is the total standard deviation. Therefore, the Pp index can be expressed as:
U −L
Pp =
6σ t
An estimate, σˆ t , of the total standard deviation ( σ t ) is required to obtain an estimate of the Pp index. In
practice, σˆ t will be the standard deviation (St) of all of the data. When this has been obtained, the index is
estimated:
6.2.2 The Ppk index
When the distribution of individual values forms a normal distribution, the median X 50 % is equal to the mean
( µ ) . Further, X 99,865 % − X 50 % and X 50 % − X 0,135 % are each equal to 3σ t . Therefore, the Ppk index is
the smaller of the two values:
U −µ
PpkU =
3σ t
or
µ−L
PpkL =
3σ t
where Ppk is estimated by:
U−X
PˆpkU =
3σˆ t
or
X −L
PˆpkL =
3σˆ t
The lower the index, the greater the proportion of items produced out-of-specification.
26
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 22514-4:2007(E)
6.3 Process performance indices for measured data (non-normal distribution)
6.3.1 General
The approaches adopted in this clause for non-normal data is the same as that given earlier in 5.5 for
capability indices.
6.3.2 Probability paper method
From graphs similar to those given in Figure 4, estimates of the quantiles X 0,135 % and X 99,865 % can be
obtained. The estimates are denoted by Y1 and Y2 respectively and the formula becomes:
U −L
Pˆp =
Y2 − Y1
In a similar way, the Ppk formulae become:
U − Xˆ 50 %
PˆpkU =
Y − Xˆ
2 50 %
or
Xˆ 50 % − L
PˆpkL =
Xˆ 50 % − Y1
whichever gives the lower value. If the index is less than a given value, the process is deemed to have
produced an excessive proportion of items outside of the specification. The proportion nonconforming
depends upon the distribution and the value of the index. The link between the index and the proportion of
nonconforming items produced depends on the class of distributions. Care should be taken not to interpret
indices on the basis of cut-off points that have derived for the normal distribution and hence are only
applicable for that distribution.
Note that the probability paper method directly estimates fairly extreme quantiles and that this can be
inaccurate. Additionally, the estimation method using probability paper, although very simple to use, is
nevertheless somewhat crude and computational procedures are preferred (see Annex C).
6.3.3 Pearson curves method
As an alternative to using probability paper, standardized Pearson curves are sometimes used. The method is
described by way of an example (see Annex B). The index is computed using the formula:
U −L
Pˆp =
Xˆ 99,865 % − Xˆ 0,135 %
where X̂ 0,135 % and X̂ 99,865 % are the estimated 0,135 % and 99,865 % quantiles from the standardized
Pearson curves.
Also, we have the formulae:
U − Xˆ 50 %
PˆpkU =
Xˆ 99,865 % − Xˆ 50 %
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
and
Xˆ 50% − L
PˆpkL =
Xˆ 50% − Xˆ 0,135%
where X̂ 50 % is the estimated median value.
In order to use this method, the user will need to establish skewness and kurtosis values in addition to the
mean and standard deviation for the data set upon which the index is to be computed.
This method is not preferred but is presented here due to its occasional use for completeness. See 5.5.3 for
further comments about the use of this method.
6.3.4 The distribution identification method
See Annex C for a description of certain families of distribution functions such as the log-normal distribution,
the Rayleigh and the Weibull distributions that are commonly found when investigating process performance.
See also 5.5.4 for further comments about the method.
6.4 Other performance indices for measured data
All of the indices given earlier for capability will have counterparts when considering performance. Any
standard deviation will represent the total variation (σ t ) instead of the inherent variation (σ ) . See 4.3 and
5.2.1.
6.5 Assessment of proportion out-of-specification (normal distribution)
The same method to estimate the proportion of out-of-specification is used here as in 5.8. The reader should
substitute PˆpkU and PˆpkL respectively for Cˆ pkU and Cˆ pkL . Table 3 may also be used to determine the
proportion out-of-specification and the reader should enter the table with PˆpkU or PˆpkL instead of Cˆ pkU or
Cˆ pkL .
6.6 Attributes
The performance indices will be the same as those given earlier. See 5.9 for a description of the measures.
--`,,```,,,,````-`-`,,`,,`,`,,`---
28
ISO/TR 22514-4:2007(E)
Annex A
(informative)
Estimating standard deviations
A.1 General
It is necessary to estimate the standard deviation in order to calculate the indices referred to in this part of
ISO 22514. There are two types of standard deviation to consider. The first is what might be described as the
short-term standard deviation or instantaneous (inherent) standard deviation. It is often calculated from
statistics taken from a control chart and this is shown in A.2. The other is the estimation of the total standard
deviation and this is described in A.3.
A.2 Inherent standard deviation
A.2.1 Estimation using the mean range value
The inherent (process) standard deviation (the data will be taken from an “in control” control chart) may be
estimated from a range control chart using the following formula:
--`,,```,,,,````-`-`,,`,,`,`,,`---
R
σˆ =
d2
where d 2 is a control chart factor taken from Table A.1.
Table A.1 — Control chart factors for the estimation of process standard deviation
Subgroup size (n) d2 c4
2 1,128 0,797 9
3 1,693 0,886 2
4 2,059 0,921 3
5 2,326 0,940 0
6 2,534 0,951 5
7 2,704 0,959 4
8 2,847 0,965 0
9 2,970 0,969 3
10 a 3,078 0,972 7
a Values for d2 and c4 may be found in textbooks for sample sizes
greater than 10.

ISO/TR 22514-4:2007(E)
A.2.2 Estimation using the mean standard deviation value
If a standard deviation control chart has been used to monitor the within subgroup variation, the inherent
(process) standard deviation can be estimated using the following formula:
S
σˆ =
c4
where c4 is a control chart factor taken from Table A.1.
A.2.3 Estimation using subgroup standard deviations
If the within subgroup standard deviation is calculated for every subgroup, the following formula gives a more
precise value than those in A.2.1 and A 2.2 for the inherent standard deviation:
m
∑S j2
j =1
σˆ =
m
where there are m subgroups of n observations in each.
A.3 Estimation of total standard deviation

When data are generated from a process which is “out-of-control” or if no control chart has been used, it is
inappropriate to compute the standard deviation using the methods of A.2. The following formula should be
used instead:
--`,,```,,,,````-`-`,,`,,`,`,,`---
N
∑( X i − X )
2
i =1
σˆ t = S t =
N −1
The circumstances that lead to the use of this equation are when fluctuations exist in the process mean
caused by assignable causes that might not be removable, and this extra variation is to be incorporated with
the remaining random cause variation. It is the appropriate measure of variation for use in calculating the
performance indices.
When considering multiple stream processes, such as a multiple cavity injection moulding press, it is often
desired to treat the data from all the cavities as if they came from a single process. The data from each cavity
might form a single normal distribution. However, the reality is often that each cavity produces a slightly
different distribution because either the means or the variabilities or both are different. If the data from all of
the process streams can be considered to yield a normal distribution, the best estimate of the process
variation will be given by this equation.
30
ISO/TR 22514-4:2007(E)
Annex B
(informative)
Estimating capability and performance measures using Pearson

curves — Procedure and example 1)
B.1 Record specification limits
Upper limit, U = 0,30
Lower limit, L = 0,20
B.2 Record process statistics

The process is shown to be statistically “in control”
Mean, x = 0,235
--`,,```,,,,````-`-`,,`,,`,`,,`---
Standard deviation, σˆ = 0,012 2
Skewness, γˆ1 = 0,7 (rounding to one decimal place)
Kurtosis, βˆ 2 = 3,5 (rounding to one decimal place)
B.3 Look up standardized 0,135 percentile

For positive skewness, use Table B.1; for negative skewness, use Table B.2.
0,135 percentile, P0,135% = 3,056 by interpolation
B.4 Look up standardized 99,865 percentile

For positive skewness, use Table B.2; for negative skewness, use Table B.1.
99,865 percentile, P99,865% = 4,656 by interpolation
B.5 Look up standardized median in Table B.3

For positive skewness, reverse the sign; for negative skewness, leave positive.
Standardized median, P50% = −0,067 5 by interpolation
1) Procedure based on that in Reference [10] (see the Bibliography).

ISO/TR 22514-4:2007(E)
B.6 Compute estimated 0,135 % quantile
Xˆ 0,135 % = x − σˆ P0,135 %
= 0,235 − ( 0,012 2 × 3,056 )
= 0,197 7
B.7 Compute estimated 99,865 % quantile
X 99,865 % = x + σˆ P99,865 %
= 0,235 + ( 0,012 2 × 4,656 )
= 0,291 8
B.8 Compute estimated median
Xˆ 50 % = x + σˆ P50 %
= 0,235 + ( 0,012 2 × −0,067 5 )
= 0,234 2
B.9 Compute process capability indices
U −L
Cˆ p =
Xˆ 99,865 % − Xˆ 0,135 %
0,30 − 0,20
=
0,291 8 − 0,197 7
= 1,06
U − Xˆ 50 %
Cˆ pkU =
Xˆ 99,865 % − Xˆ 50 %
0,30 − 0,234 2
=
0,291 8 − 0,234 2
= 1,14
Xˆ 50 % − L
Cˆ pkL =
Xˆ 50 % − Xˆ 0,135 %
0,234 2 − 0,20
=
0,234 2 − 0,197 7
= 0,94
--`,,```,,,,````-`-`,,`,,`,`,,`---
32
ISO/TR 22514-4:2007(E)
Table B.1
PEARSON CURVES (STANDARDIZED TAILS)
P0,135 % (0,135 percentile) for γ1 W 0. P99,865 % (99,865 percentile) for γ1 < 0.
Kurtosis Skewness (γ1)

(β2)
(β2) 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8 1,9 2,0
−1,4 1,512 1,421 1,317 1,206 1,092 0,979 0,868 0,762 −1,4
−1,2 1,727 1,619 1,496 1,364 1,230 1,100 0,975 0,858 0,747 −1,2
−1,0 1,966 1,840 1,696 1,541 1,384 1,232 1,089 0,957 0,836 −1,0
−0,8 2,210 2,072 1,912 1,736 1,555 1,377 1,212 1,062 0,927 0,804 0,692 −0,8
−0,6 2,442 2,298 2,129 1,941 1,740 1,539 1,348 1,175 1,023 0,887 0,766 0,656 −0,6
−0,4 2,653 2,506 2,335 2,141 1,930 1,711 1,496 1,299 1,125 0,974 0,841 0,723 0,616 −0,4
−0,2 2,839 2,692 2,522 2,329 2,116 1,887 1,655 1,434 1,235 1,065 0,919 0,791 0,677 0,574 −0,2
0,0 3,000 2,856 2,689 2,500 2,289 2,059 1,817 1,578 1,356 1,163 1,000 0,861 0,739 0,630 0,531 0,0
0,2 3,140 2,986 2,834 2,653 2,447 2,220 1,976 1,726 1,485 1,269 1,086 0,933 0,801 0,686 0,583 0,2
0,4 3,261 3,088 2,952 2,785 2,589 2,368 2,127 1,873 1,619 1,382 1,178 1,008 0,865 0,742 0,634 0,536 0,4
0,6 3,366 3,164 3,045 2,896 2,714 2,502 2,267 2,015 1,754 1,502 1,277 1,087 0,931 0,799 0,658 0,583 0,489 0,6
0,8 3,458 3,222 3,118 2,986 2,821 2,622 2,396 2,148 1,887 1,625 1,381 1,172 1,000 0,857 0,736 0,629 0,533 0,8
1,0 3,539 3,266 3,174 3,058 2,910 2,727 2,512 2,271 2,013 1,748 1,491 1,262 1,072 0,917 0,787 0,675 0,575 0,484 1,0
1,2 3,611 3,300 3,218 3,115 2,983 2,817 2,616 2,385 2,132 1,876 1,602 1,357 1,149 0,979 0,840 0,721 0,617 0,524 1,2
1,4 3,674 3,327 3,254 3,161 3,043 2,893 2,708 2,488 2,243 1,981 1,713 1,456 1,230 1,045 0,894 0,768 0,659 0,562 0,475 1,4
1,6 3,731 3,349 3,282 3,199 3,092 2,957 2,787 2,581 2,345 2,089 1,821 1,556 1,316 1,113 0,950 0,815 0,701 0,600 0,510 1,6
1,8 3,782 3,367 3,306 3,229 3,133 3,011 2,855 2,664 2,438 2,189 1,925 1,664 1,404 1,185 1,008 0,863 0,743 0,638 0,546 0,461 1,8
2,0 3,828 3,382 3,325 3,255 3,167 3,055 2,914 2,736 2,524 2,283 2,023 1,755 1,494 1,261 1,068 0,913 0,785 0,676 0,580 0,494 2,0
2,2 3,870 3,395 3,342 3,277 3,196 3,093 2,964 2,800 2,600 2,369 2,116 1,850 1,584 1,339 1,132 0,964 0,828 0,714 0,615 0,526 0,445 2,2
2,4 3,908 3,405 3,356 3,295 3,220 3,126 3,006 2,855 2,669 2,448 2,202 1,940 1,673 1,420 1,198 1,018 0,873 0,752 0,649 0,557 0,475 2,4
2,6 3,943 3,415 3,367 3,311 3,241 3,153 3,043 2,904 2,730 2,521 2,283 2,026 1,760 1,501 1,267 1,073 0,918 0,791 0,683 0,589 0,504 2,6
2,8 3,975 3,423 3,378 3,324 3,259 3,177 3,075 2,946 2,784 2,586 2,358 2,107 1,844 1,581 1,338 1,131 0,965 0,830 0,717 0,620 0,533 2,8
3,0 4,004 3,430 3,387 3,326 3,274 3,198 3,103 2,983 2,831 2,646 2,427 2,183 1,924 1,661 1,410 1,191 1,013 0,870 0,752 0,651 0,562 3,0
3,2 4,031 3,436 3,395 3,346 3,288 3,216 3,127 3,015 2,874 2,699 2,491 2,254 2,000 1,738 1,483 1,253 1,063 0,911 0,787 0,681 0,590 3,2
3,4 4,056 3,441 3,402 3,356 3,300 3,233 3,149 3,043 2,911 2,747 2,549 2,321 2,072 1,813 1,555 1,317 1,115 0,953 0,822 0,712 0,618 3,4
3,6 4,079 3,446 3,408 3,364 3,311 3,247 3,168 3,069 2,945 2,790 2,602 2,383 2,140 1,884 1,626 1,381 1,169 0,996 0,858 0,744 0,646 3,6
3,8 4,101 3,450 3,414 3,371 3,321 3,259 3,184 3,091 2,974 2,829 2,651 2,440 2,205 1,953 1,695 1,446 1,224 1,041 0,895 0,775 0,674 3,8
4,0 4,121 3,454 3,419 3,378 3,329 3,271 3,200 3,111 3,001 2,864 2,695 2,494 2,265 2,018 1,762 1,510 1,281 1,088 0,932 0,807 0,702 4,0
4,2 4,140 3,458 3,423 3,384 3,337 3,281 3,213 3,129 3,025 2,895 2,735 2,543 2,321 2,080 1,827 1,574 1,338 1,135 0,971 0,839 0,730 4,2
4,4 4,157 3,461 3,428 3,389 3,344 3,290 3,225 3,145 3,047 2,923 2,771 2,588 2,374 2,138 1,889 1,636 1,396 1,184 1,011 0,872 0,758 4,4
4,6 4,174 3,464 3,431 3,394 3,350 3,299 3,236 3,160 3,066 2,949 2,805 2,629 2,424 2,194 1,948 1,697 1,453 1,234 1,052 0,905 0,786 4,6
4,8 4,189 3,466 3,435 3,399 3,356 3,306 3,246 3,173 3,084 2,972 2,835 2,668 2,470 2,246 2,005 1,756 1,510 1,285 1,094 0,939 0,815 4,8
5,0 4,204 3,469 3,438 3,403 3,362 3,313 3,256 3,186 3,100 2,994 2,863 2,703 2,513 2,296 2,059 1,813 1,566 1,336 1,137 0,975 0,844 5,0
--`,,```,,,,````-`-`,,`,,`,`,,`---
5,2 4,218 3,471 3,441 3,406 3,367 3,320 3,264 3,197 3,114 3,013 2,888 2,735 2,562 2,342 2,111 1,867 1,621 1,387 1,181 1,010 0,874 5,2
5,4 4,231 3,473 3,444 3,410 3,371 3,326 3,272 3,207 3,128 3,031 2,911 2,765 2,589 2,386 2,160 1,920 1,675 1,438 1,225 1,047 0,904 5,4
ISO/TR 22514-4:2007(E)
Table B.1 (continued)


(β2)
(β2) 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8 1,9 2,0
5,6 4,243 3,475 3,446 3,413 3,375 3,331 3,279 3,216 3,140 3,047 2,933 2,793 2,624 2,427 2,206 1,970 1,727 1,489 1,270 1,085 0,935 5,6
5,8 4,255 3,477 3,448 3,416 3,379 3,336 3,286 3,225 3,152 3,062 2,952 2,818 2,656 2,465 2,260 2,019 1,778 1,539 1,316 1,123 0,966 5,8
6,0 4,266 3,478 3,451 3,419 3,383 3,341 3,292 3,233 3,162 3,076 2,970 2,841 2,685 2,501 2,292 2,065 1,827 1,588 1,361 1,162 0,999 6,0
6,2 4,276 3,480 3,453 3,422 3,386 3,345 3,297 3,240 3,172 3,089 2,987 2,863 2,713 2,535 2,332 2,109 1,874 1,635 1,407 1,202 1,031 6,2
6,4 4,286 3,481 3,454 3,424 3,389 3,349 3,303 3,247 3,181 3,100 3,003 2,883 2,739 2,567 2,369 2,151 1,919 1,682 1,452 1,242 1,065 6,4
6,6 4,296 3,483 3,456 3,426 3,392 3,353 3,308 3,254 3,189 3,111 3,017 2,902 2,763 2,597 2,405 2,191 1,962 1,727 1,496 1,282 1,099 6,6
6,8 4,305 3,484 3,458 3,429 3,395 3,357 3,312 3,260 3,197 3,122 3,030 2,919 2,785 2,624 2,438 2,229 2,004 1,771 1,540 1,323 1,134 6,8
7,0 4,313 3,485 3,459 3,431 3,398 3,360 3,316 3,265 3,204 3,131 3,043 2,936 2,806 2,651 2,469 2,265 2,044 1,814 1,583 1,363 1,169 7,0
7,2 4,322 3,486 3,461 3,432 3,400 3,363 3,321 3,270 3,211 3,140 3,054 2,951 2,825 2,675 2,499 2,300 2,083 1,855 1,625 1,403 1,204 7,2
7,4 4,330 3,487 3,462 3,434 3,403 3,366 3,324 3,275 3,218 3,148 3,065 2,965 2,843 2,698 2,527 2,333 2,120 1,895 1,666 1,443 1,240 7,4
--`,,```,,,,````-`-`,,`,,`,`,,`---
7,6 4,337 3,488 3,464 3,436 3,405 3,369 3,328 3,280 3,224 3,156 3,075 2,978 2,860 2,720 2,554 2,364 2,155 1,933 1,706 1,482 1,276 7,6
7,8 4,344 3,489 3,465 3,437 3,407 3,372 3,331 3,284 3,229 3,164 3,085 2,990 2,876 2,740 2,579 2,394 2,189 1,970 1,744 1,521 1,311 7,8
8,0 4,351 3,490 3,466 3,439 3,409 3,374 3,335 3,289 3,235 3,171 3,094 3,002 2,891 2,759 2,603 2,422 2,221 2,005 1,782 1,559 1,347 8,0
8,2 4,358 3,491 3,467 3,440 3,411 3,377 3,338 3,292 3,240 3,177 3,103 3,013 2,906 2,777 2,625 2,449 2,252 2,040 1,818 1,596 1,382 8,2
8,4 4,365 3,492 3,468 3,442 3,412 3,379 3,340 3,296 3,244 3,183 3,111 3,023 2,919 2,794 2,646 2,475 2,282 2,073 1,854 1,632 1,418 8,4
8,6 4,371 3,492 3,469 3,443 3,414 3,381 3,343 3,300 3,249 3,189 3,118 3,033 2,932 2,810 2,666 2,499 2,310 2,104 1,888 1,667 1,452 8,6
8,8 4,377 3,493 3,470 3,444 3,416 3,383 3,346 3,303 3,253 3,195 3,125 3,042 2,943 2,825 2,685 2,522 2,337 2,135 1,921 1,702 1,486 8,8
9,0 4,382 3,494 3,471 3,445 3,417 3,385 3,348 3,306 3,257 3,200 3,132 3,051 2,955 2,839 2,703 2,544 2,363 2,164 1,953 1,736 1,520 9,0
9,2 4,388 3,495 3,472 3,447 3,418 3,387 3,351 3,309 3,261 3,205 1,138 3,059 2,965 2,853 2,720 2,565 2,388 2,192 1,984 1,768 1,553 9,2
9,4 4,393 3,495 3,473 3,448 3,420 3,388 3,353 3,312 3,265 3,209 3,144 3,067 2,975 2,866 2,738 2,585 2,411 2,219 2,014 1,800 1,586 9,4
9,6 4,398 3,496 3,473 3,449 3,421 3,390 3,355 3,315 3,268 3,214 3,150 3,075 2,985 2,878 2,752 2,604 2,434 2,245 2,042 1,831 1,617 9,6
9,8 4,403 3,496 4,474 3,450 3,422 3,392 3,357 3,317 3,272 3,218 3,156 3,082 2,994 2,890 2,766 2,622 2,456 2,271 2,070 1,861 1,648 9,8
10,0 4,408 3,497 3,475 3,451 3,424 3,393 3,359 3,320 3,275 3,222 3,161 3,088 3,003 2,901 2,780 2,639 2,476 2,295 2,097 1,890 1,679 10,0
10,2 3,425 3,395 3,361 3,322 3,278 3,226 3,166 3,095 3,011 2,911 2,793 2,655 2,496 2,318 2,123 1,918 1,708 10,2
10,4 3,396 3,363 3,325 3,281 3,230 3,171 3,101 3,019 2,921 2,806 2,671 2,515 2,340 2,148 1,945 1,737 10,4
10,6 3,364 3,327 3,283 3,233 3,175 3,107 3,026 2,930 2,818 2,686 2,533 2,361 2,172 1,972 1,765 10,6
10,8 3,329 3,286 3,237 3,179 3,112 3,033 2,940 2,829 2,700 2,551 2,382 2,196 1,998 1,793 10,8
11,0 3,289 3,240 3,184 3,118 3,040 2,948 2,840 2,714 2,567 2,401 2,218 2,023 1,819 11,0
11,2 3,243 3,188 3,123 3,046 2,956 2,851 2,727 2,583 2,420 2,240 2,047 1,845 11,2
11,4 3,191 3,128 3,053 2,964 2,861 2,739 2,598 2,438 2,261 2,070 1,870 11,4
11,6 3,195 3,132 3,058 2,972 2,870 2,751 2,613 2,456 2,281 2,093 1,895 11,6
11,8 3,137 3,064 2,979 2,879 2,762 2,627 2,473 2,301 2,115 1,919 11,8
12,0 3,141 3,070 2,986 2,888 2,773 2,641 2,489 2,320 2,136 1,942 12,0
12,2 3,075 2,993 2,896 2,784 2,653 2,505 2,338 2,157 1,965 12,2
34
ISO/TR 22514-4:2007(E)
Table B.2

(β2)
(β2) 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8 1,9 2,0
−1,4 1,512 1,584 1,632 1,655 1,653 1,626 1,579 1,516 −1,4
−1,2 1,727 1,813 1,871 1,899 1,895 1,861 1,803 1,726 1,636 −1,2
−1,0 1,966 2,065 2,134 2,170 2,169 2,131 2,061 1,966 1,856 −1,0
−0,8 2,210 2,320 2,400 2,446 2,454 2,422 2,349 2,241 2,108 1,965 1,822 −0,8
--`,,```,,,,````-`-`,,`,,`,`,,`---
−0,6 2,442 2,560 2,648 2,704 2,726 2,708 2,646 2,540 2,395 2,225 2,052 1,885 −0,6
−0,4 2,653 2,774 2,869 2,934 2,969 2,968 2,926 2,837 2,699 2,518 2,314 2,114 1,928 −0,4
−0,2 2,839 2,961 3,060 3,133 3,179 3,194 3,173 3,109 2,993 2,824 2,608 2,373 2,152 1,952 −0,2
0,0 3,000 3,123 3,224 3,303 3,358 3,387 3,385 3,345 3,259 3,116 2,914 2,665 2,405 2,169 1,960 0,0
0,2 3,140 3,261 3,364 3,447 3,510 3,550 3,564 3,546 3,488 3,378 3,206 2,970 2,690 2,412 2,167 0,2
0,4 3,261 3,381 3,484 3,570 3,639 3,688 3,715 3,715 3,681 3,603 3,468 3,264 2,993 2,687 2,398 2,149 0,4
0,6 3,366 3,485 3,588 3,676 3,749 3,805 3,843 3,858 3,844 3,793 3,693 3,529 3,290 2,984 2,658 2,366 2,119 0,6
0,8 3,458 3,575 3,678 3,768 3,844 3,905 3,951 3,978 3,981 3,953 3,883 3,758 3,561 3,283 2,945 2,609 2,322 0,8
1,0 3,539 3,654 3,757 3,847 3,926 3,991 4,044 4,080 4,096 4,087 4,043 3,952 3,797 3,561 3,243 2,881 2,547 2,269 1,0
1,2 3,611 3,724 3,826 3,917 3,997 4,066 4,124 4,167 4,194 4,208 4,177 4,115 3,998 3,808 3,529 3,172 2,798 2,476 1,2
1,4 3,674 3,786 3,887 3,978 4,060 4,131 4,193 4,243 4,278 4,296 4,290 4,252 4,168 4,020 3,789 3,463 3,075 2,705 2,399 1,4
1,6 3,731 3,842 3,942 4,033 4,115 4,189 4,253 4,308 4,351 4,378 4,386 4,367 4,311 4,200 4,015 3,736 3,364 2,961 2,609 1,6
1,8 3,782 3,891 3,990 4,081 4,164 4,239 4,307 4,365 4,414 4,449 4,468 4,472 4,431 4,352 4,209 3,979 3,646 3,238 2,840 2,511 1,8
2,0 3,828 3,936 4,034 4,125 4,208 4,285 4,354 4,416 4,468 4,511 4,539 4,549 4,532 4,479 4,372 4,189 3,907 3,522 3,095 2,719 2,0
2,2 3,870 3,976 4,073 4,164 4,248 4,325 4,396 4,460 4,517 4,564 4,600 4,620 4,619 4,587 4,510 4,369 4,137 3,796 3,370 2,949 2,603 2,2
2,4 3,908 4,013 4,109 4,199 4,283 4,361 4,433 4,500 4,559 4,611 4,653 4,682 4,693 4,678 4,627 4,521 4,336 4,047 3,648 3,201 2,808 2,4
2,6 3,943 4,046 4,142 4,231 4,315 4,394 4,467 4,535 4,597 4,653 4,700 4,736 4,757 4,756 4,725 4,649 4,506 4,269 3,916 3,471 3,033 2,6
2,8 3,975 4,077 4,172 4,261 4,344 4,423 4,498 4,567 4,631 4,690 4,741 4,783 4,812 4,824 4,809 4,758 4,650 4,460 4,160 3,745 3,280 2,8
3,0 4,004 4,105 4,199 4,287 4,371 4,450 4,525 4,596 4,662 4,723 4,777 4,824 4,860 4,882 4,881 4,850 4,771 4,623 4,376 4,007 3,544 3,0
3,2 4,031 4,131 4,224 4,312 4,396 4,475 4,550 4,622 4,689 4,752 4,810 4,861 4,903 4,932 4,944 4,929 4,875 4,762 4,563 4,247 3,813 3,2
3,4 4,056 4,155 4,247 4,335 4,418 4,498 4,573 4,645 4,714 4,779 4,839 4,893 4,940 4,976 4,997 4,996 4,963 4,880 4,723 4,461 4,072 3,4
3,6 4,079 4,177 4,269 4,356 4,439 4,518 4,594 4,667 4,737 4,803 4,865 4,922 4,973 5,015 5,044 5,055 5,038 4,980 4,859 4,647 4,311 3,6
3,8 4,101 4,197 4,288 4,375 4,458 4,537 4,614 4,687 4,757 4,825 4,888 4,948 5,002 5,049 5,085 5,106 5,103 5,066 4,976 4,806 4,524 3,8
4,0 4,121 4,217 4,307 4,393 4,476 4,555 4,631 4,705 4,776 4,845 4,910 4,972 5,029 5,080 5,122 5,150 5,159 5,139 5,075 4,943 4,712 4,0
4,2 4,140 4,234 4,324 4,410 4,492 4,571 4,648 4,722 4,794 4,863 4,929 4,993 5,052 5,107 5,153 5,189 5,208 5,202 5,159 5,059 4,873 4,2
4,4 4,157 4,251 4,340 4,425 4,508 4,587 4,663 4,737 4,809 4,879 4,947 5,012 5,074 5,131 5,181 5,223 5,250 5,257 5,232 5,159 5,012 4,4
4,6 4,174 4,267 4,355 4,440 4,522 4,601 4,677 4,752 4,824 4,895 4,963 5,029 5,093 5,152 5,207 5,253 5,288 5,305 5,295 5,244 5,131 4,6
4,8 4,189 4,281 4,369 4,454 4,535 4,614 4,691 4,765 4,838 4,909 4,978 5,045 5,110 5,172 5,229 5,280 5,321 5,346 5,349 5,318 5,233 4,8
5,0 4,204 4,295 4,383 4,467 4,548 4,627 4,703 4,778 4,851 4,922 4,992 5,060 5,126 5,190 5,249 5,303 5,350 5,383 5,396 5,381 5,320 5,0
5,2 4,218 4,308 4,395 4,479 4,560 4,638 4,715 4,789 4,862 4,934 5,004 5,073 5,141 5,206 5,267 5,325 5,376 5,415 5,437 5,436 5,395 5,2
5,4 4,231 4,321 4,407 4,490 4,571 4,649 4,725 4,800 4,873 4,945 5,016 5,086 5,154 5,220 5,284 5,344 5,399 5,443 5,474 5,483 5,460 5,4
ISO/TR 22514-4:2007(E)


(β2)
(β2) 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8 1,9 2,0
5,6 4,243 4,332 4,418 4,501 4,581 4,659 4,736 4,810 4,884 4,956 5,027 5,097 5,166 5,233 5,299 5,361 5,418 5,468 5,505 5,525 5,516 5,6
5,8 4,255 4,343 4,429 4,511 4,591 4,669 4,745 4,820 4,893 4,966 5,037 5,108 5,177 5,246 5,312 5,376 5,436 5,491 5,533 5,561 5,565 5,8
6,0 4,266 4,354 4,439 4,521 4,600 4,678 4,754 4,829 4,902 4,975 5,046 5,117 5,188 5,257 5,325 5,390 5,452 5,511 5,558 5,593 5,608 6,0
6,2 4,276 4,364 4,448 4,530 4,609 4,695 4,763 4,837 4,911 4,983 5,055 5,126 5,197 5,267 5,336 5,403 5,467 5,529 5,581 5,621 5,645 6,2
6,4 4,286 4,373 4,457 4,538 4,618 4,703 4,771 4,845 4,919 4,991 5,063 5,135 5,206 5,276 5,346 5,414 5,480 5,542 5,600 5,646 5,678 6,4
6,6 4,296 4,382 4,466 4,547 4,626 4,710 4,778 4,853 4,926 4,999 5,071 5,143 5,214 5,285 5,356 5,425 5,492 5,557 5,618 5,669 5,706 6,6
6,8 4,305 4,391 4,474 4,554 4,633 4,717 4,785 4,860 4,933 5,006 5,078 5,150 5,222 5,293 5,364 5,434 5,503 5,569 5,634 5,688 5,732 6,8
7,0 4,313 4,399 4,481 4,562 4,640 4,724 4,792 4,867 4,940 5,013 5,085 5,157 5,229 5,301 5,372 5,443 5,513 5,581 5,648 5,706 5,754 7,0
7,2 4,322 4,406 4,489 4,569 4,647 4,730 4,799 4,873 4,946 5,019 5,091 5,164 5,236 5,308 5,380 5,451 5,522 5,591 5,658 5,722 5,775 7,2
7,4 4,330 4,414 4,496 4,576 4,654 4,736 4,805 4,879 4,952 5,025 5,097 5,170 5,242 5,314 5,387 5,459 5,530 5,601 5,669 5,736 5,792 7,4
7,6 4,337 4,421 4,503 4,582 4,660 4,742 4,811 4,885 4,958 5,031 5,103 5,175 5,248 5,320 5,393 5,466 5,538 5,609 5,679 5,749 5,808 7,6
7,8 4,344 4,428 4,509 4,588 4,666 4,747 4,817 4,890 4,963 5,036 5,109 5,181 5,253 5,326 5,399 5,472 5,545 5,617 5,688 5,760 5,823 7,8
8,0 4,351 4,434 4,515 4,594 4,672 4,753 4,822 4,896 4,969 5,041 5,114 5,186 5,259 5,331 5,404 5,478 5,551 5,624 5,696 5,771 5,836 8,0
8,2 4,358 4,441 4,521 4,600 4,677 4,758 4,827 4,901 4,974 5,046 5,118 5,191 5,263 5,336 5,410 5,483 5,557 5,631 5,704 5,775 5,847 8,2
8,4 4,365 4,447 4,527 4,605 4,682 4,762 4,832 4,905 4,978 5,051 5,123 5,195 5,268 5,341 5,414 5,488 5,562 5,637 5,710 5,783 5,858 8,4
8,6 4,371 4,452 4,532 4,611 4,687 4,767 4,837 4,910 4,983 5,055 5,127 5,200 5,272 5,345 5,419 5,493 5,587 5,642 5,717 5,790 5,867 8,6
8,8 4,377 4,458 4,538 4,616 4,692 4,772 4,841 4,914 4,987 5,059 5,132 5,204 5,276 5,349 5,423 5,497 5,572 5,647 5,722 5,797 5,875 8,8
9,0 4,382 4,463 4,543 4,621 4,697 4,776 4,845 4,918 4,991 5,063 5,135 5,208 5,280 5,353 5,427 5,501 5,576 5,652 5,727 5,803 5,883 9,0
9,2 4,388 4,468 4,548 4,625 4,701 4,780 4,850 4,923 4,995 5,067 5,139 5,211 5,284 5,357 5,431 5,505 5,580 5,656 5,732 5,808 5,890 9,2
9,4 4,393 4,473 4,552 4,630 4,705 4,784 4,854 4,926 4,999 5,071 5,143 5,215 5,287 5,361 5,434 5,509 5,584 5,660 5,736 5,813 5,889 9,4
9,6 4,398 4,478 4,557 4,634 4,710 4,788 4,857 4,930 5,002 5,074 5,146 5,218 5,291 5,364 5,437 5,512 5,587 5,663 5,740 5,817 5,894 9,6
9,8 4,403 4,483 4,561 4,638 4,714 4,791 4,861 4,934 5,006 5,078 5,149 5,222 5,294 5,367 5,440 5,515 5,590 5,667 5,744 5,821 5,898 9,8
10,0 4,408 4,487 4,565 4,642 4,717 4,795 4,865 4,937 5,009 5,081 5,153 5,225 5,297 5,370 5,443 5,518 5,593 5,670 5,747 5,825 5,903 10,0
10,2 4,721 4,798 4,868 4,940 5,012 5,084 5,156 5,228 5,300 5,373 5,446 5,521 5,596 5,673 5,750 5,828 5,906 10,2
10,4 4,871 4,943 5,015 5,087 5,158 5,230 5,303 5,375 5,449 5,523 5,599 5,675 5,753 5,831 5,910 10,4
10,6 4,874 4,947 5,018 5,090 5,161 5,233 5,305 5,378 5,451 5,526 5,601 5,678 5,755 5,834 5,913 10,6
10,8 4,949 5,021 5,092 5,164 5,236 5,308 5,380 5,454 5,528 5,603 5,680 5,757 5,836 5,915 10,8
--`,,```,,,,````-`-`,,`,,`,`,,`---
11,0 5,024 5,095 5,166 5,238 5,310 5,383 5,456 5,530 5,605 5,682 5,760 5,838 5,918 11,0
11,2 5,098 5,169 5,240 5,312 5,385 5,458 5,532 5,607 5,684 5,762 5,840 5,920 11,2
11,4 5,171 5,243 5,314 5,387 5,460 5,534 5,609 5,686 5,763 5,842 5,922 11,4
11,6 5,173 5,245 5,316 5,389 5,462 5,536 5,611 5,687 5,765 5,844 5,924 11,6
11,8 5,247 5,318 5,391 5,464 5,538 5,613 5,689 5,767 5,845 5,925 11,8
12,0 5,249 5,320 5,393 5,465 5,539 5,614 5,690 5,768 5,847 5,927 12,0
12,2 5,322 5,394 5,467 5,541 5,616 5,692 5,769 5,848 5,928 12,2
36
ISO/TR 22514-4:2007(E)
Table B.3
PEARSON CURVES (STANDARDIZED MEDIAN)
P50 % (50 percentile). Change sign for γ1 > 0.

(β2)
(β2) 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8 1,9 2,0
−1,4 0,000 0,053 0,111 0,184 0,282 0,424 0,627 0,754 −1,4
--`,,```,,,,````-`-`,,`,,`,`,,`---
−1,2 0,000 0,039 0,082 0,132 0,196 0,284 0,412 0,591 0,727 −1,2
−1,0 0,000 0,031 0,065 0,103 0,151 0,212 0,297 0,419 0,586 −1,0
−0,8 0,000 0,026 0,054 0,085 0,123 0,169 0,231 0,317 0,439 0,598 0,681 −0,8
−0,6 0,000 0,023 0,047 0,073 0,104 0,142 0,190 0,254 0,343 0,468 0,616 0,653 −0,6
−0,4 0,000 0,020 0,041 0,064 0,091 0,122 0,161 0,212 0,280 0,375 0,504 0,633 0,616 −0,4
−0,2 0,000 0,018 0,037 0,058 0,081 0,108 0,141 0,183 0,237 0,311 0,413 0,542 0,638 0,574 −0,2
0,0 0,000 0,017 0,034 0,053 0,073 0,097 0,126 0,161 0,206 0,266 0,347 0,456 0,579 0,621 0,531 0,0
0,2 0,000 0,015 0,032 0,049 0,068 0,089 0,114 0,145 0,183 0,233 0,299 0,388 0,501 0,605 0,582 0,2
0,4 0,000 0,014 0,029 0,045 0,063 0,082 0,105 0,132 0,165 0,208 0,263 0,336 0,433 0,545 0,607 0,536 0,4
0,6 0,000 0,013 0,028 0,043 0,059 0,077 0,097 0,122 0,151 0,188 0,235 0,297 0,379 0,481 0,579 0,579 0,489 0,6
0,8 0,000 0,013 0,026 0,040 0,055 0,072 0,091 0,113 0,140 0,172 0,213 0,266 0,336 0,425 0,527 0,590 0,533 0,8
1,0 0,000 0,012 0,025 0,038 0,053 0,068 0,086 0,106 0,130 0,159 0,196 0,242 0,301 0,379 0,474 0,563 0,569 0,484 1,0
1,2 0,000 0,011 0,024 0,036 0,050 0,065 0,082 0,100 0,122 0,148 0,181 0,222 0,274 0,341 0,426 0,520 0,576 0,524 1,2
1,4 0,000 0,011 0,023 0,035 0,048 0,062 0,078 0,095 0,116 0,140 0,169 0,206 0,252 0,310 0,385 0,474 0,554 0,555 0,475 1,4
1,6 0,000 0,010 0,022 0,034 0,046 0,060 0,074 0,091 0,110 0,132 0,159 0,192 0,233 0,285 0,351 0,432 0,518 0,564 0,510 1,6
1,8 0,000 0,010 0,021 0,032 0,044 0,057 0,072 0,087 0,105 0,126 0,151 0,180 0,217 0,264 0,323 0,396 0,480 0,549 0,540 0,461 1,8
2,0 0,000 0,009 0,020 0,031 0,043 0,055 0,069 0,084 0,101 0,120 0,143 0,171 0,204 0,246 0,299 0,365 0,443 0,521 0,552 0,494 2,0
2,2 0,000 0,009 0,020 0,030 0,042 0,054 0,067 0,081 0,097 0,115 0,137 0,162 0,193 0,231 0,279 0,338 0,410 0,488 0,544 0,522 0,445 2,2
2,4 0,000 0,009 0,019 0,029 0,040 0,052 0,065 0,078 0,094 0,111 0,131 0,155 0,183 0,218 0,261 0,315 0,381 0,456 0,524 0,538 0,475 2,4
2,6 0,000 0,008 0,018 0,029 0,039 0,051 0,063 0,076 0,091 0,107 0,126 0,148 0,175 0,207 0,246 0,295 0,355 0,426 0,498 0,539 0,503 2,6
2,8 0,000 0,008 0,018 0,028 0,038 0,049 0,061 0,074 0,088 0,104 0,122 0,143 0,167 0,197 0,233 0,278 0,333 0,398 0,470 0,526 0,522 2,8
3,0 0,000 0,008 0,017 0,027 0,037 0,048 0,059 0,072 0,085 0,101 0,118 0,138 0,161 0,189 0,222 0,263 0,313 0,374 0,443 0,506 0,530 3,0
3,2 0,000 0,008 0,017 0,027 0,037 0,047 0,058 0,070 0,083 0,098 0,114 0,133 0,155 0,181 0,212 0,250 0,296 0,352 0,417 0,483 0,525 3,2
3,4 0,000 0,008 0,017 0,026 0,036 0,046 0,057 0,068 0,081 0,095 0,111 0,129 0,150 0,174 0,203 0,239 0,281 0,333 0,394 0,460 0,513 3,4
3,6 0,000 0,007 0,016 0,025 0,035 0,045 0,056 0,067 0,079 0,093 0,108 0,125 0,145 0,168 0,196 0,228 0,268 0,316 0,373 0,437 0,495 3,6
3,8 0,000 0,007 0,016 0,025 0,034 0,044 0,054 0,066 0,078 0,091 0,105 0,122 0,141 0,163 0,188 0,219 0,256 0,301 0,354 0,415 0,475 3,8
4,0 0,000 0,007 0,015 0,025 0,034 0,043 0,053 0,064 0,076 0,089 0,103 0,119 0,137 0,158 0,182 0,211 0,246 0,288 0,337 0,395 0,455 4,0
4,2 0,000 0,007 0,015 0,024 0,033 0,043 0,053 0,063 0,075 0,087 0,101 0,116 0,133 0,153 0,176 0,204 0,236 0,276 0,322 0,376 0,435 4,2
4,4 0,000 0,007 0,015 0,024 0,033 0,042 0,052 0,062 0,073 0,085 0,099 0,113 0,130 0,149 0,171 0,197 0,228 0,265 0,308 0,359 0,416 4,4
4,6 0,000 0,007 0,015 0,023 0,032 0,041 0,051 0,061 0,072 0,084 0,097 0,111 0,127 0,145 0,167 0,191 0,220 0,255 0,296 0,344 0,399 4,6
4,8 0,000 0,006 0,015 0,023 0,032 0,041 0,050 0,060 0,071 0,082 0,095 0,109 0,124 0,142 0,162 0,186 0,213 0,246 0,285 0,330 0,382 4,8
5,0 0,000 0,006 0,014 0,023 0,031 0,040 0,049 0,059 0,070 0,081 0,093 0,107 0,122 0,139 0,158 0,181 0,207 0,238 0,274 0,317 0,367 5,0
5,2 0,000 0,006 0,014 0,022 0,031 0,040 0,049 0,058 0,069 0,080 0,092 0,105 0,119 0,136 0,155 0,176 0,201 0,231 0,265 0,306 0,353 5,2
5,4 0,000 0,006 0,014 0,022 0,030 0,039 0,048 0,057 0,068 0,078 0,090 0,103 0,117 0,133 0,151 0,172 0,196 0,224 0,257 0,295 0,340 5,4

ISO/TR 22514-4:2007(E)

PEARSON CURVES (STANDARDIZED MEDIAN)
P50 % (50 percentile). Change sign for γ1 > 0.

(β2)
(β2) 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8 1,9 2,0
5,6 0,000 0,006 0,014 0,022 0,030 0,039 0,047 0,057 0,067 0,077 0,089 0,101 0,115 0,131 0,148 0,168 0,191 0,218 0,249 0,285 0,328 5,6
5,8 0,000 0,006 0,014 0,022 0,030 0,038 0,047 0,056 0,066 0,076 0,087 0,100 0,113 0,128 0,145 0,164 0,186 0,212 0,242 0,277 0,317 5,8
6,0 0,000 0,006 0,014 0,021 0,029 0,038 0,046 0,055 0,065 0,075 0,086 0,098 0,111 0,126 0,142 0,161 0,182 0,207 0,235 0,268 0,307 6,0
6,2 0,000 0,006 0,013 0,021 0,029 0,037 0,046 0,055 0,064 0,074 0,085 0,097 0,110 0,124 0,140 0,158 0,178 0,202 0,229 0,261 0,298 6,2
6,4 0,000 0,006 0,013 0,021 0,029 0,037 0,045 0,054 0,063 0,073 0,084 0,096 0,108 0,122 0,137 0,155 0,175 0,197 0,223 0,254 0,289 6,4
6,6 0,000 0,006 0,013 0,021 0,028 0,037 0,045 0,054 0,063 0,073 0,083 0,094 0,107 0,120 0,135 0,152 0,171 0,193 0,218 0,247 0,281 6,6
6,8 0,000 0,006 0,013 0,021 0,028 0,036 0,044 0,053 0,062 0,072 0,082 0,093 0,105 0,118 0,133 0,150 0,168 0,189 0,213 0,241 0,273 6,8
7,0 0,000 0,005 0,013 0,020 0,028 0,036 0,044 0,053 0,061 0,071 0,081 0,092 0,104 0,117 0,131 0,147 0,165 0,185 0,209 0,236 0,267 7,0
7,2 0,000 0,005 0,013 0,020 0,028 0,036 0,044 0,052 0,061 0,070 0,080 0,091 0,103 0,115 0,129 0,145 0,162 0,182 0,205 0,230 0,260 7,2
7,4 0,000 0,005 0,013 0,020 0,027 0,035 0,043 0,052 0,060 0,070 0,079 0,090 0,101 0,114 0,128 0,143 0,160 0,179 0,201 0,226 0,254 7,4
7,6 0,000 0,005 0,012 0,020 0,027 0,035 0,043 0,051 0,060 0,069 0,079 0,089 0,100 0,113 0,126 0,141 0,157 0,176 0,197 0,221 0,249 7,6
7,8 0,000 0,005 0,012 0,020 0,027 0,035 0,043 0,051 0,059 0,068 0,078 0,088 0,099 0,111 0,124 0,139 0,155 0,173 0,193 0,217 0,243 7,8
8,0 0,000 0,005 0,012 0,019 0,027 0,034 0,042 0,050 0,059 0,068 0,077 0,087 0,098 0,110 0,123 0,137 0,153 0,170 0,190 0,213 0,238 8,0
8,2 0,000 0,005 0,012 0,019 0,027 0,034 0,042 0,050 0,058 0,067 0,076 0,086 0,097 0,109 0,121 0,135 0,151 0,168 0,187 0,209 0,234 8,2
8,4 0,000 0,005 0,012 0,019 0,026 0,034 0,042 0,050 0,058 0,067 0,076 0,086 0,096 0,108 0,120 0,134 0,149 0,165 0,184 0,205 0,229 8,4
8,6 0,000 0,005 0,012 0,019 0,026 0,034 0,041 0,049 0,057 0,066 0,075 0,085 0,095 0,107 0,119 0,132 0,147 0,163 0,181 0,202 0,225 8,6
8,8 0,000 0,005 0,012 0,019 0,026 0,033 0,041 0,049 0,057 0,066 0,075 0,084 0,094 0,106 0,118 0,131 0,145 0,161 0,179 0,199 0,221 8,8
9,0 0,000 0,005 0,012 0,019 0,026 0,033 0,041 0,049 0,057 0,065 0,074 0,084 0,094 0,105 0,116 0,129 0,143 0,159 0,176 0,196 0,218 9,0
9,2 0,000 0,005 0,012 0,019 0,026 0,033 0,040 0,048 0,056 0,065 0,073 0,083 0,093 0,104 0,115 0,128 0,142 0,157 0,174 0,193 0,214 9,2
9,4 0,000 0,005 0,012 0,019 0,026 0,033 0,040 0,048 0,056 0,064 0,073 0,082 0,092 0,103 0,114 0,127 0,140 0,155 0,172 0,190 0,211 9,4
9,6 0,000 0,005 0,012 0,019 0,025 0,033 0,040 0,048 0,055 0,064 0,072 0,082 0,091 0,102 0,113 0,125 0,139 0,153 0,170 0,188 0,208 9,6
9,8 0,000 0,005 0,012 0,018 0,025 0,032 0,040 0,047 0,055 0,063 0,072 0,081 0,091 0,101 0,112 0,124 0,137 0,152 0,168 0,185 0,205 9,8
10,0 0,000 0,005 0,011 0,018 0,025 0,032 0,040 0,047 0,055 0,063 0,071 0,080 0,090 0,100 0,111 0,123 0,136 0,150 0,166 0,183 0,202 10,0
10,2 0,000 0,032 0,039 0,047 0,054 0,063 0,071 0,080 0,089 0,099 0,110 0,122 0,135 0,149 0,164 0,181 0,200 10,2
10,4 0,000 0,032 0,039 0,047 0,054 0,062 0,071 0,079 0,089 0,099 0,109 0,121 0,133 0,147 0,162 0,179 0,197 10,4
10,6 0,000 0,039 0,046 0,054 0,062 0,070 0,079 0,088 0,098 0,109 0,120 0,132 0,146 0,160 0,177 0,195 10,6
10,8 0,000 0,046 0,054 0,061 0,070 0,078 0,088 0,097 0,108 0,119 0,131 0,144 0,159 0,175 0,192 10,8
11,0 0,000 0,053 0,061 0,069 0,078 0,087 0,097 0,107 0,118 0,130 0,143 0,157 0,173 0,190 11,0
11,2 0,000 0,061 0,069 0,078 0,087 0,096 0,106 0,117 0,129 0,142 0,156 0,171 0,188 11,2
11,4 0,000 0,069 0,077 0,086 0,095 0,105 0,116 0,128 0,141 0,154 0,169 0,186 11,4
11,6 0,000 0,068 0,077 0,086 0,095 0,104 0,116 0,127 0,139 0,153 0,168 0,184 11,6
11,8 0,000 0,076 0,085 0,094 0,104 0,115 0,126 0,138 0,152 0,166 0,182 11,8
12,0 0,000 0,076 0,085 0,094 0,104 0,114 0,125 0,137 0,150 0,165 0,181 12,0
12,2 0,000 0,084 0,093 0,103 0,113 0,124 0,136 0,149 0,163 0,179 12,2
--`,,```,,,,````-`-`,,`,,`,`,,`---
38
ISO/TR 22514-4:2007(E)
Annex C
(informative)
Distribution identification
C.1 General
Sometimes, the form of the distribution is known or can be reasonably assumed and can be verified by tests-
of-fit. The approach is to estimate the parameters of that distribution and to use them to derive the relevant
quantiles from which the capability estimates are obtained. The proportion out-of-specification can be directly
estimated.
The method is illustrated below with some often encountered distributions.
C.2 Normal distribution

2 2
If X1, ..., XN is a sample from a normal distribution with mean µ and variance σ , the estimates of µ and σ are
N
1
µˆ = X =
N i =1
∑
Xi
and
N
1
∑( )
2
σˆ 2 = Xi − X
N − 1 i =1
The capability indices can in turn be estimated using the formulae given in this part of ISO 22514. Thus:
U −L
Cˆ p =
6σˆ
U − µˆ
Cˆ pkU =
3σˆ
µ
ˆ −L
Cˆ pkL =
3σˆ
and, finally,
(
Cˆ pk = min Cˆ pkL , Cˆ pkU )
The estimated proportion out-of-specification items below L is calculated as:
pˆ L = 1 − Φ 3Cˆ pkL( )
and the estimated proportion out-of-specification items above U is calculated as:
pˆ U = 1 − Φ 3Cˆ pkU( )
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
here Φ denotes the distribution function of the standard normal distribution. The actual calculations of pˆ L and
pˆ U may be performed as outlined in 5.8.
C.3 Log-normal distribution
C.3.1 General
The log-normal distribution with parameters µ and σ has probability density function
1 ⎛ (log X − µ ) 2 ⎞
f ( x) = exp ⎜ − ⎟
X σ 2π ⎜ 2σ 2 ⎟
⎝ ⎠
where X > 0, and log denotes the natural logarithm, i.e. the logarithm to base e. When X has a log-normal
distribution with parameters µ and σ, then logX has a normal distribution with mean µ and variance σ2.
--`,,```,,,,````-`-`,,`,,`,`,,`---
If X1, ..., XN is a sample from a log-normal distribution, the data can be transformed to normality by taking the
logarithms logX1, ..., logXN. The calculations in C.2 can then be used. Alternatively, the calculations can be
made directly in the original scale of the measurements. The two methods are given below. In both cases, the
parameters are estimated as:
N
1
µˆ = X = ∑
N i =1
Xi
and
N
1
∑( )
2
σˆ 2 = Xi − X
N − 1 i =1
C.3.2 Log-normal distribution — Transformation to normality
The upper and lower specification limits must be transformed so they become logU and logL. The formulae in
C.2 can then be applied. For example, the estimates of Cp, CpkL, and CpkU are calculated as:
logU − log L
Cˆ p =
6σˆ
log U − µˆ
Cˆ pkU =
3σˆ
µˆ − log L
Cˆ pkL =
3σˆ
CpkL and CpkU may in turn be inserted in the appropriate formulae in C.2 to give the estimated proportions out-
of-specification.
C.3.3 Log-normal distribution — Original scale
The quantiles of the log-normal distribution are:
X α = exp(σΦ −1(α ) + µ )
40
ISO/TR 22514-4:2007(E)
where Φ−1 is the inverse of the distribution function of the standard normal distribution. In particular:
X 0,135% = e −3σ + µ
X 50% = e µ
X 99,865% = e 3σ + µ
and the estimated indices are:

U −L
Cˆ p =
3σˆ + µˆ
e − e −3σˆ + µˆ
U − e µˆ
Cˆ pkU =
e 3σˆ + µˆ − e µˆ
e µˆ − L
Cˆ pkL =
e µˆ − e −3σˆ + µˆ
These indices will be numerically different from those obtained using the transformation approach of C.3.2. A
process owner with log-normally distributed items may well develop a “feel” for the indices, but their
interpretation should not routinely be based on the cut-off points used for indices calculated for normal data.
The estimated proportion of out-of-specification items is calculated using the specification limits and the
distribution function of the log-normal distribution. Thus:
⎛ log L − µˆ ⎞
pˆ L = Φ ⎜ ⎟
⎝ σˆ ⎠
and
⎛ logU − µˆ ⎞
pˆ U = 1 − Φ ⎜ ⎟
⎝ σˆ ⎠
These estimates are exactly the same as those obtained with the transformation approach in C.3.2.
C.4 Rayleigh distribution

This distribution is used almost exclusively to describe position, eccentricity and run-off in two-dimensional
problems. In these situations, it is typical for there to be only a one-sided specification limit, U.
The Rayleigh distribution has distribution function
⎛ X2 ⎞
F ( x ) = 1 − exp ⎜ − ⎟
⎜ 2θ 2 ⎟
⎝ ⎠
where X > 0 and θ is a positive parameter. If X1, ..., XN is a sample from a Rayleigh distribution, the parameter
θ is estimated by:
N
∑ X i2
i =1
θˆ =
2N
41
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 22514-4:2007(E)
and the estimated proportion out-of-specification is given by:
⎛ ⎞
⎜ ⎟
⎜ NU 2 ⎟
pˆ U = exp ⎜ − N ⎟
⎜ 2⎟
⎜ ∑
Xi ⎟
⎝ i =1 ⎠
C.5 Weibull distribution

This is a distribution of great versatility. It is often used in analysing data gathered during reliability studies
when the patterns that describe the data are often irregular and non-normal. The distribution has three
parameters:
1) ξ a scale parameter;
--`,,```,,,,````-`-`,,`,,`,`,,`---
2) β a shape parameter; and
3) γ a location parameter, frequently assumed to be 0.
There are occasions when a capability study will not produce data that follow a normal distribution pattern, but
the Weibull distribution can be used to describe the data and thereafter provide a way to calculate capability or
performance measures.
The distribution function of the Weibull distribution is:
⎛ ⎛ X −γ β⎞
⎞
F ( X ) = 1 − exp ⎜ − ⎜ ⎟ ⎟
⎜ ⎝ ξ ⎠ ⎟
⎝ ⎠
and hence the quantiles of Weibull distribution are:
−β
(
X α % = Y + ξ − log (1 − p ) )
In particular, the quantiles X0,135 %, X50 %, and X99,865 % and subsequently, the capability indices can be
calculated. More importantly the proportions out-of-specification items are:
⎛ ⎛ L −γ β⎞
⎞
p L = F ( L ) = 1 − exp ⎜ − ⎜ ⎟ ⎟
⎜ ⎝ ξ ⎠ ⎟
⎝ ⎠
and
⎛ ⎛U −γ β⎞
⎞
pU = 1 − F (U ) = exp ⎜ − ⎜ ⎟ ⎟
⎜ ⎝ ξ ⎠ ⎟
⎝ ⎠
The estimated proportions out-of-specification are obtained using estimates of the parameters.
C.6 Folded normal distribution

The folded normal distribution, sometimes called the half-normal distribution, is often used to describe the
variation in a characteristic whose specification includes geometrical tolerances. This situation gives a one-
sided specification. It typically applies where geometrical characteristics, form and orientation are specified.
42
ISO/TR 22514-4:2007(E)
The folded normal distribution, with parameters µ and σ, has probability density function:
⎧ 2⎫
2 ⎪ 1⎡ X −µ ⎤ ⎪
f ( x) = exp ⎨ − ⎢ ⎥ ⎬
σ 2π ⎪⎩ 2 ⎣⎢ σ ⎥⎦ ⎪⎭
where 0 u X < ∞.
The folded normal distribution is proportional to the normal distribution. The distribution’s estimated
proportions can therefore be found using a standard normal distribution table with the appropriate table value
multiplied by 2.
C.7 Other distributions

The distributions given above are those frequently to be encountered. There exist many others that the reader
will find documented in the numerous text books on statistics.
--`,,```,,,,````-`-`,,`,,`,`,,`---
ISO/TR 22514-4:2007(E)
Annex D
(informative)
Confidence intervals
D.1 Normal distribution
D.1.1 General
It is important for any person having calculated a capability index to realize that the computed value is only an
estimate of the true value of that index. Usually, the greater the amount of data used to compute the index the
better the estimate will be. The following paragraphs are intended to alert the reader to this fact and to provide
methods that confidence intervals for the indices can be calculated.
The confidence interval calculations are only appropriate when the mean and not the median has been the
chosen measure of location.
D.1.2 Normal distribution — Formula method
The 1 − α confidence intervals are:
Cˆ p
Cˆ p ± z1−α 2
2N − 2
2
1 Cˆ pk U
Cˆ pkU ± z1−α 2 +
9N 2N − 2
2
1 Cˆ pk L
Cˆ pkL ± z1−α 2 +
9 N 2N − 2
where z is the standardized variate for the normal distribution. These calculations should be done with at least
50 readings.
NOTE These formulae assume the estimated indices have been computed using a standard deviation based on the
total sample size (N).
D.1.3 Normal distribution — Tabular method for the Cp index
D.1.3.1 General
An example of the method to estimate a confidence interval for a Cp index is shown in D.1.3.2.
The method requires multipliers Kl and Ku are read from Table D.1. These differ according to how many
readings the estimated Cp has been based on. The index is multiplied by these multipliers to give the
confidence interval. This calculation should be done with at least 50 readings.
As an example, the 95 % confidence interval would be:
K l95% Cˆ p u Cp u K u 95% Cˆ p
--`,,```,,,,````-`-`,,`,,`,`,,`---
44
ISO/TR 22514-4:2007(E)
D.1.3.2 Procedure and example
a) Record estimated Cp value and total sample size:
Cˆ p = 1,20 and N = 100
b) Select the required confidence level:
Confidence level = 95 %
--`,,```,,,,````-`-`,,`,,`,`,,`---
c) Read multipliers from Table D.1:
K l 95 % = 0,86
Ku = 1,14
95 %
d) Calculate confidence interval:
K l95 % Cˆ p u Cp u K u 95 % Cˆ p
0,86 × 1,20 u C p u 1,14 × 1,20
1,03 u C p u 1,37
Table D.1 — Multipliers for the confidence interval of the Cp index
Total sample size, N

Confidence interval Multiplier
50 75 100 150 300
Kl 0,83 0,86 0,88 0,90 0,93
90 %
Ku 1,16 1,13 1,12 1,09 1,07
Kl 0,80 0,84 0,86 0,89 0,92

95 %
Ku 1,20 1,16 1,14 1,11 1,08
Kl 0,75 0,79 0,82 0,85 0,90

99 %
Ku 1,26 1,21 1,18 1,15 1,11
D.2 Other confidence intervals

Confidence intervals for other calculated indices (not graphical estimates) can be found for non-normal
distributions as well as for normal distributions.

ISO/TR 22514-4:2007(E)
Annex E
(informative)
Example of the output from computer software
This example is illustrated using a software program called MINITAB™.2)
Measurement
5 4 5 4 6 8 4 4 4 6 9 6 5 4 4
7 11 3 3 5 5 5 6 6 6 5 7 5 7 8
4 7 7 6 4 5 6 4 7 5 5 3 8 9 5
10 6 2 6 3
Table of Percentiles
Standard
Distribution Percent Percentiles Error 95.0% CI
Largest Extreme Value 0.135 1.79066 0.326324 1.2 2.4
Standard
Largest Extreme Value 50 5.28275 0.257354 4.8 5.8
Standard
Largest Extreme Value 99.865 14.9478 1.19875 12.6 17.3
The probability plot is shown in Figure E.1.

--`,,```,,,,````-`-`,,`,,`,`,,`---
2) MINITAB is the trade name of a product supplied by Minitab Inc. This information is given for the convenience of users
of this document and does not constitute an endorsement by ISO of the product named. Equivalent products may be used
if they can be shown to lead to the same results.
46
ISO/TR 22514-4:2007(E)
Figure E.1
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
Bibliography
[1] ISO 3534-1:2006, Statistics — Vocabulary and symbols — Part 1: General statistical terms and terms
used in probability
[2] ISO 3534-2:2006, Statistics — Vocabulary and symbols — Part 2: Applied statistics
[3] ISO 5479, Statistical interpretation of data — Tests for departure from the normal distribution
[4] ISO 7870-1, Control charts — Part 1: General guidelines
--`,,```,,,,````-`-`,,`,,`,`,,`---
[5] ISO 8258, Shewhart control charts
[6] ISO 9000:2005, Quality management systems — Fundamentals and vocabulary
[7] ISO/IEC 12207:1995, Information technology — Software life cycle processes
[8] STEPHENS, M.A. Anderson-Darling Test for Goodness of Fit. In: Encyclopedia of Statistical Sciences.
Vol. 1, pp. 81-85, (eds. Johnson, N.L. and Kotz, S.), Wiley Interscience, 1982, ISBN 0-471-05546-8
[9] KOTZ, S. and LOVELACE, C.R. Process Capability Indices in Theory and Practice. Hodder Arnold, 1998
[10] CLEMENTS, J.A. Process capability calculations for non-normal distributions. Quality Process, 22
(1989), pp. 95-100
48
--`,,```,,,,````-`-`,,`,,`,`,,`---

ISO/TR 22514-4:2007(E)
ICS 03.120.30
Price based on 48 pages
--`,,```,,,,````-`-`,,`,,`,`,,`---


Standard ISO TR 22514 4 2007 9658 PDF

Uploaded by

Copyright:

Available Formats

Standard ISO TR 22514 4 2007 9658 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Standard ISO TR 22514 4 2007 9658 PDF

Uploaded by

Copyright:

Available Formats

TECHNICAL ISO/TR

Statistical methods in process

Copyright International Organization for Standardization © ISO 2007

COPYRIGHT PROTECTED DOCUMENT

Annex D (informative) Confidence intervals...................................................................................................44

© ISO 2007 – All rights reserved

⎯ Part 1: General principles and concepts

⎯ Part 3: Machine performance studies for measured data on discrete parts

⎯ Part 4: Process capability estimates and performance measures [Technical Report]

© ISO 2007 – All rights reserved

Copyright International Organization for Standardization

Statistical methods in process management — Capability and

2 Terms and definitions

2.1 Basic terms

NOTE 1 A characteristic can be inherent or assigned.

NOTE 2 A characteristic can be qualitative or quantitative.

NOTE 3 There are various classes of characteristics, such as the following:

⎯ physical (e.g. mechanical, electrical, chemical or biological characteristics);

⎯ sensory (e.g. relating to smell, touch, taste, sight, hearing);

⎯ behavioural (e.g. courtesy, honesty, veracity);

⎯ temporal (e.g. punctuality, reliability, availability);

⎯ ergonomic (e.g. physiological characteristic, or related to human safety);

⎯ functional (e.g. maximum speed of an aircraft).

[ISO 9000:2005, 3.5.1]

© ISO 2007 – All rights reserved

NOTE 1 Inherent means existing in something, especially as a permanent characteristic.

[ISO 9000:2005, 3.5.2]

[ISO 3534-2:2006, 3.1.3]

[ISO 3534-2:2006, 3.1.6]

[ISO 3534-2:2006, 3.1.2]

[ISO 3534-2:2006, 2.5.1]

EXAMPLE 3 The unimodal continuous distributions.

NOTE 3 Adapted from ISO 3534-2:2006, 2.5.2.

[ISO 3534-2:2006, 2.5.3]

[ISO 3534-2:2006, 2.5.7]

© ISO 2007 – All rights reserved

EXAMPLE In a normal distribution (2.1.7), with mean, µ, and standard deviation σ :

pU is the upper fraction nonconforming;

Φ is the distribution function of the standard normal distribution;

U is the upper specification limit.

[ISO 3534-2:2006, 2.5.4]

EXAMPLE In a normal distribution (2.1.7), with mean, µ, and standard deviation, σ :

pL is the lower fraction nonconforming;

Φ is the distribution function of the standard normal distribution;

L is the lower specification limit.

[ISO 3534-2:2006, 2.5.5]

EXAMPLE In a normal distribution (2.1.7), with mean, µ, and standard deviation, σ :

pt is the total fraction nonconforming;

Φ is the distribution function of the standard normal distribution;

U is the upper specification limit;

L is the lower specification limit.

[ISO 3534-2:2006, 2.5.6]

2.2 Process capability, estimates and indices

R is the average range calculated from a set of m subgroup ranges;

Sj is the sample standard deviation of the jth subgroup;

m is the number of subgroups of the same size, n;

d2, c4 are constants based on subgroup size, n.

and X j is the observed mean of the jth subgroup.

⎯ becomes progressively less efficient as the subgroup size increases;