This action might not be possible to undo. Are you sure you want to continue?
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 1 OF 82 EDITION 2  JANUARY 2007
M3003 EDITION 2  JANUARY 2007
The Expression of Uncertainty and Confidence in
Measurement JH 2007
CONTENTS
SECTION PAGE
1 Introduction 2
2 Overview 4
3 More detail 10
4 Type A evaluation of standard uncertainty 18
5 Type B evaluation of standard uncertainty 20
6 Reporting of results 21
7 Step by step procedure for evaluation of measurement uncertainty 22
Appendix A Best measurement capability 26
Appendix B Deriving a coverage factor for unreliable input quantities 28
Appendix C Dominant nonGaussian Type B uncertainty 31
Appendix D Derivation of the mathematical model 34
Appendix E Some sources of error and uncertainty in electrical calibrations 37
Appendix F Some sources of error and uncertainty in mass calibrations 42
Appendix G Some sources of error and uncertainty in temperature calibrations 44
Appendix H Some sources of error and uncertainty in dimensional calibrations 45
Appendix J Some sources of error and uncertainty in pressure calibration using DWTs 46
Appendix K Examples of application for calibration 48
Appendix L Expression of uncertainty for a range of values 66
Appendix M Assessment of compliance with specification 69
Appendix N Uncertainties for test results 74
Appendix P Electronic data processing 78
Appendix Q Symbols 80
Appendix R References 82
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 2OF 82
1 INTRODUCTION
1.1 The general requirements that testing and calibration laboratories have to meet if they wish to
demonstrate that they operate to a quality system, are technically competent and are able to
generate technically valid results are contained within ISO/IEC 17025:2005. This international
standard forms the basis for international laboratory accreditation and in cases of differences in
interpretation remains the authoritative document at all times. M3003 is not intended as a
prescriptive document, and does not set out to introduce additional requirements to those in
ISO/IEC 17025:2005 but to provide amplification and guidance on the current requirements within
the international standard.
1.2 The purpose of these guidelines is to provide policy on the evaluation and reporting of
measurement uncertainty for testing and calibration laboratories. Related topics, such as
evaluation of compliance with specifications, are also included. A number of worked examples are
included in order to illustrate how practical implementation of the principles involved can be
achieved.
1.3 The guidance in this document is based on information in the Guide to the Expression of
Uncertainty in Measurement
[1]
, hereinafter referred to as the GUM. M3003 is consistent with the
GUM both in methodology and terminology. It does not, however, preclude the use of other
methods of uncertainty evaluation that may be more appropriate to a specific discipline. For
example, the use of Bayesian statistics is becoming recognised as being particularly useful in
certain areas of testing.
1.4 M3003 is aimed both at the beginner and at those more experienced in the subject of
measurement uncertainty. In order to address the needs of an audience with a wide spectrum of
experience, the subject is introduced in relatively straightforward terms and gives details of the
basic concepts involved. Crossreferences are made to a number of Appendices, where more
detailed information is presented for those who wish to obtain a deeper understanding of the
subject.
1.5 Edition 2 of M3003 is a complete revision of the previous issue and it is impractical to list all the
changes in detail. Some of the more notable changes or additions are as follows:
1.5.1 An overview of the basic concepts relating to uncertainty evaluation is given in Section 2 to
introduce these ideas to those new to the subject. This is then expanded upon in Section 3, which
gives more formal detail.
1.5.2 ISO/IEC 17025:2005 criteria, which were not in place when Edition 1 was published, have been
considered.
1.5.3 A section on derivation of the measurement model has been included.
1.5.4 Concepts are accompanied by simple worked examples as they are introduced.
1.5.5 A number of diagrams illustrating the concepts have been included.
1.5.6 Further detail has been included regarding dominant contributions.
1.5.7 The subject of compliance with specification has been explored in more detail.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 3OF 82
1.5.8 The worked examples have been reviewed and minor amendments have been made as
necessary.
1.5.9 A new example, relating to pressure calibration using a deadweight tester, has been added,
together with a section describing the sources of uncertainty for these measurements.
1.5.10 Hyperlinks have been included between various sections of the document.
1.5.11 Analytical concepts such as Monte Carlo simulation and Bayesian statistics have been introduced.
1.6 No further changes of any significance will be made to M3003 Edition 2 during its lifetime.
However, minor text modifications of an editorial nature may be made if the need is identified. Any
such changes will be listed below.
Date Details of amendment
2 January 2007 First issue of M3003 Edition 2
8 January 2007 Amendment A1: Minor text corrections made; paragraph numbers in
Appendix M corrected; crossreference on Page 16 corrected.
30 January 2007 Amendment A2: Probability distribution in second line of uncertainty budget
K8.15 corrected; reference to LAB14 on page 82 updated.
22 February 2008 Amendment A3: Sensitivity coefficient C
S
in K7.6 corrected from 0.200 ºC/μV
to C
S
= 0.189 ºC/μV. Resolution in K8.14 corrected from 20 Pa to 200 Pa.
Repeated line removed in K8.1.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 4OF 82
2 OVERVIEW
2.1 In many aspects of everyday life, we are accustomed to the doubt that arises when estimating how
large or small things are. For example, if somebody asks, “what do you think the temperature of
this room is?” we might say, “it is about 23 degrees Celsius”. The use of the word “about” implies
that we know the room is not exactly 23 degrees, but is somewhere near it. In other words, we
recognise that there is some doubt about the value of the temperature that we have estimated.
2.2 We could, of course, be a bit more specific. We could say, “it is 23 degrees Celsius give or take a
couple of degrees”. The term “give or take” implies that there is still doubt about the estimate, but
now we are assigning limits to the extent of the doubt. We have given some quantitative
information about the doubt, or uncertainty, of our estimate.
2.3 It is also quite reasonable to assume that we may be more sure that our estimate is within, say, 5
degrees of the “true” room temperature than we are that it is within 2 degrees. The larger the
uncertainty we assign, the more confident we are that it encompasses the “true” value. Hence, for a
given situation, the uncertainty is related to the level of confidence.
2.4 So far, our estimate of the room temperature has been based on a subjective evaluation. This is
not entirely a guess, as we may have experience of exposure to similar and known environments.
However, in order to make a more objective measurement it is necessary to make use of a
measuring instrument of some kind; in this case we can use a thermometer.
2.5 Even if we use a measuring instrument, there will still be some doubt, or uncertainty, about the
result. For example we could ask:
“Is the thermometer accurate?”
“How well can I read it?”
“Is the reading changing?”
“I am holding the thermometer in my hand. Am I warming it up?”
“The relative humidity in the room can vary considerably. Will this affect my results?”
“Does it matter where in the room I take the measurement?”
All these factors, and possibly others, may contribute to the uncertainty of our measurement of the
room temperature.
2.6 In order to quantify the uncertainty of the room temperature measurement we will therefore have to
consider all the factors that could influence the result. We will have to make estimates of the
possible variations associated with these influences. Let us consider the questions posed above.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 5OF 82
2.7 Is the thermometer accurate?
2.7.1 In order to find out, it will be necessary to compare it with a thermometer whose accuracy is better
known. This thermometer, in turn, will have to be compared with an even better characterised one,
and so on. This leads to the concept of traceability of measurements, whereby measurements at all
levels can be traced back to agreed references. In most cases, ISO/IEC17025:2005 requires that
measurements are traceable to SI units. This is usually achieved by an unbroken chain of
comparisons to a national metrology institute, which maintains measurement standards that are
directly related to SI units.
In other words, we need a traceable calibration. This calibration itself will provide a source of
uncertainty, as the calibrating laboratory will assign a calibration uncertainty to the reported values.
When used in a subsequent evaluation of uncertainty, this is often referred to as the imported
uncertainty.
2.7.2 In terms of the thermometer accuracy, however, a traceable calibration is not the end of the story.
Measuring instruments change their characteristics as time goes by. They “drift”. This, of course, is
why regular recalibration is necessary. It is therefore important to evaluate the likely change since
the instrument was last calibrated.
If the instrument has a reliable history it may be possible to predict what the reading error will be at
a given time in the future, based on past results, and apply a correction to the reading. This
prediction will not be perfect and therefore an uncertainty on the corrected value will be present. In
other cases, the past data may not indicate a reliable trend, and a limit value may have to be
assigned for the likely change since the last calibration. This can be estimated from examination of
changes that occurred in the past. Evaluations made using these methods yield the uncertainty due
to secular stability, or changes with time, of the instrument. This is also known as “drift”.
2.7.3 There are other possible influences relating to the thermometer accuracy. For example, suppose
we have a traceable calibration, but only at 15 °C, 20 °C and 25 °C. What does this tell us about its
indication error at 23 °C?
In such cases we will have to make an estimate of the error, perhaps using interpolation between
points where calibration data is available. This is not always possible as it depends on the
measured data being such that accurate interpolation is practical. It may then be necessary to use
other information, such as the manufacturer’s specification, to evaluate the additional uncertainty
that arises when the reading is not directly at a point that has been calibrated.
2.8 How well can I read it?
2.8.1 There will inevitably be a limit to which we can resolve the reading we observe on the thermometer.
If it is a liquidinglass thermometer, this limit will often be imposed by our ability to interpolate
between the scale graduations. If it is a thermometer with a digital readout, the finite number of
digits in the display will define the limit.
2.8.2 For example, suppose the last digit of a digital thermometer can change in steps of 0.1 °C. The
reading happens to be 23.4 °C. What does this mean in terms of uncertainty?
The reading is a rounded representation of an infinite continuum of underlying values that the
thermometer would indicate if it had more digits available. In the case of a reading of 23.4 °C, this
means that the underlying value cannot be less than 23.35 °C, otherwise the rounded reading
would be 23.3 °C. Similarly, the underlying value cannot be more than 23.45 °C, otherwise the
rounded reading would be 23.5 °C.
A reading of 23.4 °C therefore means that the underlying value is somewhere between 23.35 °C
and 23.45 °C. In other words, the 0.1 °C resolution of the display has caused a rounding error
somewhere between –0.05 °C and +0.05 °C. As we have no way of knowing where in this range
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 6OF 82
the underlying value is, we have to assume the rounding error is zero with limits of ± 0.05 °C.
2.8.3 It can therefore be seen that there will always be an uncertainty of ± half of the change represented
by one increment of the last displayed digit. This does not only apply to digital displays; it applies
every time a number is recorded. If we write down a rounded result of 123.456, we are imposing an
identical effect by the fact that we have recorded this result to three decimal places, and an
uncertainty of ± 0.0005 will arise.
2.8.4 This source of uncertainty is frequently referred to as “resolution”, however it is more correctly the
numeric rounding caused by finite resolution.
2.9 Is the reading changing?
2.9.1 Yes, it probably is! Such changes may be due to variations in the room temperature itself,
variations in the performance of the thermometer and variations in other influence quantities, such
as the way we are holding the thermometer.
So what can be done about this?
2.9.2 We could, of course, just record one reading and say that it is the measured temperature at a given
moment and under particular conditions. This would have little meaning, as we know that the next
reading, a few seconds later, could well be different. So which is “correct”?
2.9.3 In practice, we will probably take an average of several measurements in order to obtain a more
realistic reading. In this way, we can “smooth out” the effect of short term variations in the
thermometer indication. The average, or arithmetic mean, of a number of readings can often be
closer to the “true” value than any individual reading is.
2.9.4 However, we can only take a finite number of measurements. This means that we will never obtain
the “true” mean value that would be revealed if we could carry out an infinite (or very large) number
of measurements. There will be an unknown error – and therefore an uncertainty – represented by
the difference from our calculated mean value and the underlying “true” mean value.
2.9.5 This uncertainty cannot be evaluated using methods like those we have already considered. Up
until now, we have looked for evidence, such as calibration uncertainty and secular stability. We
have considered what happens with finite resolution by logical reasoning. The effects of variation
between readings cannot be evaluated like this, because there is no background information
available upon which to base our evaluation.
2.9.5 The only information we have is a series of readings and a calculated average, or mean, value. We
therefore have to use a statistical approach to determine how far our calculated mean could be
away from the “true” mean. These statistics are quite straightforward and give us the uncertainty
associated with the repeatability (or, more correctly, nonrepeatability) of our measurements. This
uncertainty is referred to as the experimental standard deviation of the mean. For the sake of
brevity, this is often referred to as simply the standard deviation of the mean.
NOTE
In earlier textbooks on the subjects of uncertainty or statistics, this may be referred to as the standard error of the mean.
2.9.6 It is often convenient to regard the calculation of the standard deviation of the mean as a twostage
process, and it can be performed easily on most scientific calculators.
2.9.7 First we calculate the estimated standard deviation using the values we have measured. This
facility is indicated on most calculators by the function key xσ
n1
. On some calculators it is identified
as s(x) or simply s.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 7OF 82
2.9.8 The standard deviation of the mean is then obtained by dividing the value obtained in 2.9.7 by the
square root of the number of measurements that contributed to the mean value.
2.9.9 Let us try an example. Suppose we record five consecutive readings with our thermometer. These
are 23.0 °C, 23.4 °C, 23.1 °C, 23.6 °C and 22.9 °C.
2.9.10
Using the calculator function xσ
n1
, we obtain an estimated standard deviation of 0.2915 °C.
2.9.11 Five measurements contributed to the mean value, so we divide 0.2915 °C by the square root of 5.
236 . 2
2915 . 0
5
2915 . 0
= = 0.1304 °C.
2.9.12 Further information on the statistical analysis processes used for evaluation of nonrepeatability
can be found in Section 4.
2.10 I am holding the thermometer in my hand. Am I warming it up?
2.10.1 Quite possibly. There may be heat conduction from the hand to the temperature sensor. There may
be radiated heat fromthe body impinging on the sensor. These effects may or may not be
significant, but we will not know until an evaluation is performed. In this case, special experiments
may be required in order to determine the significance of the effect.
2.10.2 How could we do this? Some fairly basic and obvious methods come to mind. For example, we
could set up the thermometer in a temperaturestable environment and read it remotely, without the
operator nearby. We could then compare this result with that obtained when the operator is holding
it in the usual manner, or in a variety of manners. This would yield empirical data on the effects of
heat conduction and radiation. If this turns out to be significant, we could either improve the method
so that operator effects are eliminated, or we could include a contribution to measurement
uncertainty based on the results of the experiment.
2.10.3 This reveals a number of important issues. First, that the measurement may not be independent of
the operator and that special consideration may have to be given to operator effects. We may have
to train the operator to use the equipment in a particular way. Special experiments may be
necessary to evaluate particular effects. Additionally, and significantly, evaluation of uncertainty
may reveal ways in which the method can be improved, thus giving more reliable results. This is a
positive benefit of uncertainty evaluation.
2.11 The relative humidity in the room can vary considerably. Will this affect my results?
2.11.1 Maybe. If we are using a liquid in glass thermometer, it is difficult to see how the relative humidity
could significantly affect the expansion of the liquid. However, if we are using a digital thermometer
it is quite possible that relative humidity could affect the electronics that amplify and process the
signal from the sensor. The sensor itself could also be affected by relative humidity.
2.11.2 As with other influences, we need means of evaluating any such effects. In this case, we could
expose the thermometer to an environment in which the temperature can be maintained constant
but the relative humidity can be varied. This will reveal how sensitive the thermometer is to the
quantity we are concerned about.
2.11.3 This also raises a general point that is applicable to all measurements. Every measurement we
make has to be carried out in an environment of some kind; it is unavoidable. So we have to
consider whether any particular aspect of the environment could have an effect on the
measurement result.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 8OF 82
2.11.4 The significance of a particular aspect of the environment has to be considered in the light of the
specific measurement being made. It is difficult to see how, for example, gravity could significantly
influence the reading on a digital thermometer. However, it certainly will affect the results obtained
on a precision weighing machine that might be right next to the thermometer!
2.11.5 The following environmental effects are amongst the most commonly encountered when
considering measurement uncertainty:
Temperature
Relative humidity
Barometric pressure
Electric or magnetic fields
Gravity
Electrical supplies to measuring equipment
Air movement
Vibration
Light and optical reflections
Furthermore, some of these influences may have little effect as long as they remain constant, but
could affect measurement results when they start changing. Rate of change of temperature can be
particularly important.
2.11.6 It can be seen by now that understanding of a measurement system is important in order to identify
and quantify the various uncertainties that can arise in a measurement situation. Conversely,
analysis of uncertainty can often yield a deeper understanding of the system and reveal ways in
which the measurement process can be improved. This leads on to the next question…
2.12 Does it matter where in the room I make the measurement?
2.12.1 It depends what we are trying to measure! Are we interested in the temperature at a specific
location? Or the average of the temperatures encountered at any location within the room? Or the
average temperature at bench height?
2.12.2 There may be further, related questions. For example, do we require the temperature at a particular
time of day, or the average over a specific period of time?
2.12.3 Such questions have to be asked, and answered, in order that we can devise an appropriate
measurement method that gives us the information we require. Until we know the details of the
method, we are not in a position to evaluate the uncertainties that will arise from that method.
2.12.4 This leads to what is perhaps the most important question of all, one that should be asked before
we even start with our evaluation of uncertainty:
2.13 “What exactly is it that I am trying to measure?”
2.13.1 Until this question is answered, we are not in a position to carry out a proper evaluation of the
uncertainty. The particular quantity subject to measurement is known as the measurand. In order to
evaluate the uncertainty in a measurement system, we must define the measurand otherwise we
are not in a position to know how a particular influence quantity affects the value we obtain for it.
2.13.2 The implication of this is that there has to be a defined relationship between the influence quantities
and the measurand. This relationship is known as the mathematical model. This is an equation that
describes how each influence quantity affects the value assigned to the measurand. In effect, it is a
description of the measurement process. Further details about the derivation of the mathematical
model can be found in Appendix D. A proper analysis of this process also gives the answer to
another important question:
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 9OF 82
2.14 “Am I actually measuring the quantity that I thought I was measuring?”
2.14.1 Some measurement systems are such that the result would be only an approximation to the “true”
value, even if no other uncertainties were present, because of assumptions and approximations
inherent in the method. The model should include any such assumptions and therefore
uncertainties that arise from them will be accounted for in the analysis.
2.15 Summary
2.15.1 This section of M3003 has given an overview of uncertainty and some insights into how
uncertainties might arise. It has shown that we have to know our measurement system and the way
in which the various influences can affect the result. It has also shown that analysis of uncertainty
can have positive benefits in that it can reveal where enhancements can be made to measurement
methods, hence improving the reliability of measurement results.
2.15.2 The following sections of M3003 explore the issues identified in this overview in more detail.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 10 OF 82
3 IN MORE DETAIL…
3.1 The Overview section of M3003 has provided an introduction to the subject of uncertainty evaluation
and has explored a number of the issues involved. This section provides a more formal description of
these processes, using terminology consistent with that in the GUM.
3.2 A quantity (Q) is a property of a phenomenon, body or substance to which a magnitude can be
assigned. The purpose of a measurement is to assign a magnitude to the measurand; the quantity
intended to be measured. The assigned magnitude is considered to be the best estimate of the value
of the measurand.
3.3 The uncertainty evaluation process will encompass a number of influence quantities that affect the
result obtained for the measurand. These influence, or input, quantities are referred to as X and the
output quantity, i.e., the measurand, is referred to as Y.
3.4 As there will usually be several influence quantities, they are differentiated from each other by the
subscript i, So there will be several input quantities called X
i
, where i represents integer values from 1
to N, N being the number of such quantities. In other words, there will be input quantities of X
1
, X
2
, …
X
N
.
3.5 Each of these input quantities will have a corresponding value. For example, one quantity might be the
temperature of the environment – this will have a value, say 23 °C. A lowercase “x” represents the
values of the quantities. Hence the value of X
1
will be x
1
, that of X
2
will be x
2
, and so on.
3.6 The purpose of the measurement is to determine the value of the measurand, Y. As with the input
uncertainties, the value of the measurand is represented by the lowercase letter, i.e. y. The
uncertainty associated with y will comprise a combination of the input, or x
i
, uncertainties. One of the
first steps is to establish the mathematical relationship between the values of the input quantities, x
i
,
and that of the measurand, y. This process is examined in Appendix D.
3.7 The values x
i
of the input quantities X
i
will all have an associated uncertainty. This is referred to as
u(x
i
), i.e. “the uncertainty of x
i
”. These values of u(x
i
) are known as standard uncertainties – but more
on this shortly.
3.8 Some uncertainties, particularly those associated with the determination of repeatability, have to be
evaluated by statistical methods. Others have been evaluated by examining other information, such as
data in calibration certificates, evaluation of longterm drift, consideration of the effects of environment,
etc.
3.9 The GUM differentiates between statistical evaluations and those using other methods. It categorises
them into two types – Type A and Type B.
3.10 A Type A evaluation of uncertainty is carried out using statistical analysis of a series of observations.
Further details about Type A evaluations can be found in Section 4.
3.11
A Type B evaluation of uncertainty is carried out using methods other than statistical analysis of a
series of observations. Further details about Type B evaluations can be found in Section 5.
3.12 In paragraph 3.3.4 of the GUM it is stated that the purpose of the Type Aand Type B classification is to
indicate the two different ways of evaluating uncertainty components, and is for convenience in
discussion only. Whether components of uncertainty are classified as `random' or `systematic' in relation
to a specific measurement process, or described as Type A or Type B depending on the method of
evaluation, all components regardless of classification are modelled by probability distributions quantified
by variances or standard deviations.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 11 OF 82
3.13 Therefore any convention as to how they are classified does not affect the estimation of the total
uncertainty. But it should always be remembered that, in this publication, when the terms `random' and
`systematic' are used they refer to the effects of uncertainty on a specific measurement process. It is the
usual case that random components require Type Aevaluations and systematic components require
Type B evaluations, but there are exceptions.
3.14 For example, a random effect can produce a fluctuation in an instrument's indication, which is both
noiselike in character and significant in terms of uncertainty. It may then only be possible to estimate
limits to the range of indicated values. This is not a common situation but when it occurs a Type B
evaluation of the uncertainty component will be required. This is done by assigning limit values and an
associated probability distribution, as in the case of other Type B uncertainties.
3.15 The input uncertainties, associated with the values x
i
of the influence quantities X
i
, arise in a number of
forms. Some may be characterised as limit values within which little is known about the most likely
place within the limits where the “true” value may lie. A good example of this is the numeric rounding
caused by finite resolution described in paragraph 2.8. In this example, it is equally likely that the
underlying value is anywhere within the defined limits of ± half of the change represented by one
increment of the last displayed digit. This concept is illustrated in Figure 1.
3.16 a a
probability p
x
i
 a x
i
x
i
+ a
Figure 1
The expectation value x
i
lies in the centre of a distribution of possible values
with a halfwidth, or semirange, of a.
3.17 In the resolution example, a = 0.5 of a least significant digit.
3.18 It can be seen from this that there is equal probability of the value of x
i
being anywhere within the
range x
i
 a to x
i
+ a, and zero probability of it being outside these limits.
3.19 Thus, a contribution of uncertainty from the influence quantity can be characterised as a probability
distribution, i.e. a range of possible values with information about the most likely value of the input
quantity x
i
. In this example, it is not possible to say that any particular position of x
i
within the range is
more or less likely than any other. This is because there is no information available upon which to
make such a judgement.
3.20 The probability distributions associated with the input uncertainties are therefore a reflection of the
available knowledge about that particular quantity. In many cases, there will be insufficient information
available to make a reasoned judgement and therefore a uniform, or rectangular, probability
distribution has to be assumed. Figure 1 is an example of such a distribution.
3.21 If more information is available, it may be possible to assign a different probability distribution to the
value of a particular input quantity. For example, a measurement may be taken as the difference in
readings on a digital scale – typically, the zero reading will be subtracted from a reading taken further
up the scale. If the scale is linear, both of these readings will have an associated rectangular
distribution of identical size. If two identical rectangular distributions, each of magnitude ± a, are
combined then the resulting distribution will be triangular with a semirange of ± 2a.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 12 OF 82
probability p
x
i
 2a x
i
x
i
+ 2a
Figure 2
Combination of two identical rectangular distributions, each
with semirange limits of ± a, yields a triangular distribution
with a semirange of ± 2a.
3.22 There are other possible distributions that may be assigned. For example, when making
measurements of radiofrequency power an uncertainty arises due to imperfect matching between the
source and the termination. The imperfect match usually involves an unknown phase angle. This
means that a cosine function characterises the probability distribution for the uncertainty. Harris and
Warner
[3]
have shown that a symmetrical Ushaped probability distribution arises from this effect. In
this example, the distribution has been evaluated from a theoretical analysis of the principles involved.
x
i
 a x
i
x
i
+ a
Figure 3
Ushaped distribution, associated with RF mismatch uncertainty. For this
situation, x
i
is likely to be close to one or other of the edges of the
distribution.
3.23 An evaluation of the effects of nonrepeatability, performed by statistical methods, will usually yield a
Gaussian or normal distribution. Further details on this process can be found in Section 4.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 13 OF 82
3.24 When a number of distributions of whatever form are combined it can be shown that, apart from in
exceptional cases, the resulting probability distribution tends to the normal form in accordance with the
Central Limit Theorem
[4]
. The importance of this is that it makes it possible to assign a confidence level in
terms of probability to the combined uncertainty. The exceptional case arises when one contribution to
the total uncertainty dominates; in this circumstance the resulting distribution departs little from that of the
dominant contribution.
NOTE: If the dominant contribution is itself normal in form, then clearly the resulting distribution will also be normal.
Figure 4
The normal, or Gaussian, probability distribution. This is obtained when a
number of distributions, of any form, are combined and the conditions of
the Central Limit Theorem are met. In practice, if three or more distributions
of similar magnitude are present, they will combine to form a reasonable
approximation to the normal distribution. The size of the distribution is
described in terms of a standard deviation. The shaded area represents ± 1
standard deviation from the centre of the distribution. This corresponds to
approximately 68% of the area under the curve.
3.25 When the input uncertainties are combined, a normal distribution will usually be obtained. The normal
distribution is described in terms of a standard deviation. It will be therefore be necessary to express
the input uncertainties in terms that, when combined, will cause the resulting normal distribution to be
expressed at the one standard deviation level, like the example in Figure 4.
3.26 As some of the input uncertainties are expressed as limit values (e.g., the rectangular distribution),
some processing is needed to convert them into this form, which is known as a standard uncertainty
and is referred to as u(x
i
).
3.27 When it is possible to assess only the upper and lower bounds of an error, a rectangular probability
distribution should be assumed for the uncertainty associated with this error. Then, if a
i
is the semirange
limit, the standard uncertainty is given by ( )
3
i
a
i
x u = . Table 1 gives the expressions for this and for
other situations.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 14 OF 82
Table 1
Assumed
probability
distribution
Expression used to
obtain the standard
uncertainty
Comments or examples
Rectangular
( )
3
i
a
i
x u =
A digital thermometer has a least significant digit of 0.1°C.
The numeric rounding caused by finite resolution will have
semirange limits of 0.05°C. Thus the corresponding
standard uncertainty will be ( ) = = =
732 . 1
05 . 0
3
i
a
i
x u 0.029°C.
Ushaped
( )
2
i
a
i
x u =
A mismatch uncertainty associated with the calibration of an
RF power sensor has been evaluated as having semirange
limits of 1.3%. Thus the corresponding standard uncertainty
will be ( ) = = =
414 . 1
3 . 1
2
i
a
i
x u 0.92%
Triangular
( )
6
i
a
i
x u =
A tensile testing machine is used in a testing laboratory
where the air temperature can vary randomly but does not
depart from the nominal value by more than 3°C. The
machine has a large thermal mass and is therefore most
likely to be at the mean air temperature, with no probability
of being outside the 3°C limits. It is reasonable to assume a
triangular distribution, therefore the standard uncertainty for
its temperature is ( ) = = =
449 . 2
3
6
i
a
i
x u 1.2°C
Normal
(from
repeatability
evaluation)
( ) ( ) q s
i
x u =
A statistical evaluation of repeatability gives the result in
terms of one standard deviation; therefore no further
processing is required.
Normal
(from a
calibration
certificate)
( )
k
U
i
x u =
A calibration certificate normally quotes an expanded
uncertainty U at a specified, high coverage probability. A
coverage factor, k, will have been used to obtain this
expanded uncertainty from the combination of standard
uncertainties. It is therefore necessary to divide the
expanded uncertainty by the same coverage factor to obtain
the standard uncertainty.
Normal
(from a
manufacturer’s
specification)
( )
k
i
x u
limit Tolerance
=
Some manufacturers’ specifications are quoted at a given
coverage probability (sometimes referred to as confidence
level), e.g. 95% or 99%. In such cases, a normal distribution
can be assumed and the tolerance limit is divided by the
coverage factor k for the stated coverage probability. For a
coverage probability of 95%, k = 2 and for a coverage
probability of 99%, k = 2.58.
If a coverage probability is not stated then a rectangular
distribution should be assumed.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 15 OF 82
3.28 The quantities X
i
that affect the measurand Y may not have a direct, one to one, relationship with it.
Indeed, they may be entirely different units altogether. For example, a dimensional laboratory may use
steel gauge blocks for calibration of measuring tools. A significant influence quantity is temperature.
Because the gauge blocks have a significant temperature coefficient of expansion, there is an
uncertainty that arises in their length due to an uncertainty in temperature units.
3.29 In order to translate the temperature uncertainty into an uncertainty in length units, it is necessary to
know how sensitive the length of the gauge block is to temperature. In other words, a sensitivity
coefficient is required.
3.30 The sensitivity coefficient simply describes how sensitive the result is to a particular influence quantity.
In this example, the steel used in the manufacture of gauge blocks has a temperature coefficient of
expansion of approximately +11.5 x 10
 6
per °C. So, in this case, this figure can be used as the
sensitivity coefficient.
The sensitivity coefficient associated with each input estimate x
i
is referred to as c
i
. It is the partial
derivative of the model function f with respect to X
i
, evaluated at the input estimates x
i
. It is given by
3.31
In other words, it describes how the output estimate y varies with a corresponding small change in an
input estimate x
i
.
3.32 The calculations required to obtain sensitivity coefficients by partial differentiation can be a lengthy
process, particularly when there are many input contributions and uncertainty estimates are needed for a
range of values. If the functional relationship is not known for a particular measurement system the
sensitivity coefficients can sometimes be obtained by the practical approach of changing one of the input
variables by a known amount, while keeping all other inputs constant, and noting the change in the
output estimate. This approach can also be used if f is known but the determination of the partial
derivatives is likely to be difficult.
3.33 Amore straightforward approach is to replace the partial derivative ∂f / ∂x
i
by the quotient Δf / Δx
i
, where
Δf is the change in f resulti ng from a change Δx
i
in x
i
. It is important to choose the magnitude of the
change Δx
i
around x
i
carefully. It should be balanced between being sufficiently large to obtain adequate
numerical accuracy in Δf and sufficiently small to provide a mathematically sound approximation to the
partial derivative. The following example illustrates this, and why it is necessary to know the functional
relationship between the influence quantities and the measurand.
Example
The height h of a flagpole is determined
by measuring the angle obtained when
observing the top of the pole at a
specified distance d. Thus h = d tan Φ.
Both h and d are in units of length but are
related by tan Φ. In other words,
h = f(d) = d tan Φ.
If the measured distance is 7.0 m and the
measured angle is 37°, the estimated
height is 7.0 x tan (37) = 5.275 m.
Φ
d
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 16 OF 82
3.34 If the uncertainty in d is, say, ± 0.1 m then the estimate of h could be anywhere between
(7.0 – 0.1) tan (37) and (7.0 + 0.1) tan (37), i.e. between 5.200 m and 5.350 m. A change of ± 0.1 m in
the input quantity x
i
has resulted in a change of ± 0.075 m in the output estimate y. The sensitivity
coefficient is therefore 75 . 0
1 . 0
075 . 0
= .
3.35 Similar reasoning can be applied to the uncertainty in the angle Φ. If the uncertainty in Φ is ± 0.5°,
then the estimate of h could be anywhere between 7.0 tan (36.5) and 7.0 tan (37.5), i.e. between
5.179 m and 5.371 m. A change of ± 0.5°in the input quantity x
i
has resulted in a change of ± 0.096 m
in the output estimate y. The sensitivity coefficient is therefore 192 . 0
5 . 0
096 . 0
= metre per degree.
3.36 Once the standard uncertainties x
i
and the sensitivity coefficients c
i
have been evaluated, the
uncertainties have to be combined in order to give a single value of uncertainty to be associated with
the estimate y of the measurand Y. This is known as the combined standard uncertainty and is given
the symbol u
c
(y).
3.37 The combined standard uncertainty is calculated as follows:
3.38
( ) ( ) ( )
¿ ¿
= =
÷ =
N
i
i
N
i
i i c
y u x u c y u
1
2
1
2
2
(1)
3.39 In other words, the individual standard uncertainties, expressed in terms of the measurand, are
squared; these squared values are added and the square root is taken.
3.40 An example of this process is presented below, using the data from the measurement of the flagpole
height described on page 15. For the purposes of the example, it is assumed that the repeatability of
the process has been evaluated by making repeat measurements of the flagpole height, giving an
estimated standard deviation of the mean of 0.05 metres. See Section 4 for further details about the
evaluation of repeatability.
Source of uncertainty Value
Probability
distribution
Divisor
Sensitivity
coefficient
Standard uncertainty
u
i
(y)
Distance from
flagpole
0.1 m Rectangular √3 0.75
75 . 0
3
1 . 0
· = 0.0433 m
Angle measurement 0.5° Rectangular √3
0.192
metre/degree
192 . 0
3
5 . 0
· = 0.0554 m
Repeatability 0.05 m Normal 1 1 1
1
05 . 0
· = 0.05 m
Combined standard uncertainty u
c
(y) =
2 2 2
05 . 0 0554 . 0 0433 . 0 + +
= 0.0863 m
NOTE
The values in the righthand column of the table are now scaled in accordance with the effect of the corresponding input quantity
on the measurand and are expressed as a standard uncertainty.
3.41 In accordance with the Central Limit Theorem, this combined standard uncertainty takes the form of a
normal distribution. As the input uncertainties had been expressed in terms of a standard uncertainty,
the resulting normal distribution is expressed as one standard deviation, as illustrated in Figure 5.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 17 OF 82
y
5.275 m
y – 0.0863 m y + 0.0863 m
Figure 5
The measured value y is at the centre of a normal distribution
with a standard deviation equal to u
c
(y). The figures shown
relate to the example discussed above.
3.42 For a normal distribution, 1 standard deviation encompasses 68.27% of the area under the curve. This
means that there is about 68% confidence that the measured value y lies within the stated limits.
3.43
The GUM recognises the need for providing a high level of confidence – referred to herein as coverage
probability  associated with an uncertainty and uses the termexpanded uncertainty, U, which is obtained
by multiplying the combined standard uncertainty by a coverage factor. The coverage factor is given the
symbol k, thus the expanded uncertainty is given by
U= k u
c
(y). (2)
3.44 In accordance with generally accepted international practice, it is recommended that a coverage factor of
k = 2 is used to calculate the expanded uncertainty. This value of k will give a coverage probability of
approximately 95%, assuming a normal distribution.
NOTE: Acoverage factor of k = 2 actually provides a coverage probability of 95.45% for a normal distribution. For convenience this
is approximated to 95% which would relate to a coverage factor of k =1.96. However, the difference is not generally significant
since, in practice, the coverage probability is usually based on conservative assumptions and approximations to the true probability
distributions.
3.45 Example: The measurement of the height of the flagpole had a combined standard uncertainty u
c
(y) of
0.0863 m. Hence the expanded uncertainty U = k u
c
(y) = 2 x 0.0863 = 0.173 m.
3.46 There may be situations where a normal distribution cannot be assumed and a different coverage
factor may be needed in order to obtain a coverage probability of approximately 95%. Such situations
are described in Appendix B and Appendix C.
3.47 There may also be situations where a normal distribution can be assumed, but a different coverage
probability is required. For example, in safetycritical situations a higher coverage probability may be
more appropriate. The table below gives the coverage factor necessary to obtain various levels of
confidence for a normal distribution.
Coverage probability Coverage factor
p k
90% 1.64
95% 1.96
95.45% 2.00
99% 2.58
99.73% 3.00
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 18 OF 82
4 TYPE A EVALUATION OF STANDARD UNCERTAINTY
4.1 If an uncertainty is evaluated by statistical analysis of a series of observations, it is known as a
Type A evaluation.
4.2 AType Aevaluation will normally be used to obtain a value for the repeatability or randomness of a
measurement process. For some measurements, the random component of uncertainty may not be
significant in relation to other contributions to uncertainty. It is nevertheless desirable for any
measurement process that the relative importance of random effects be established. When there is a
significant spread in a sample of measurement results, the arithmetic mean or average of the results
should be calculated. If there are n independent repeated values for a quantity Q then the mean
value q is given by
n
q q q q
q
n
q
n
n
j
j
3 2 1
1
1
+ +
= =
¿
=
(3)
4.3 The spread in the results gives an indication of the repeatability of the measurement process, which
depends on various factors, including the apparatus used, the method, and sometimes on the person
making the measurement. Agood description of this spread of values is the standard deviation σ of
the n values that comprise the sample, which is given by
( )
2
1
1
¿
=
÷ =
n
j
j
q q
n
σ (4)
4.4 This expression yields the standard deviation σ of the particular set of values sampled. However,
these are not the only values that could have been sampled. If the process is repeated, another set
of values, with different values of q and σ, will be obtained.
4.5
For large values of n, these mean values approach the central limit of a distribution of all possible
values. This probability distribution can often be assumed to have the normal form.
4.6 As it is impractical to capture all values that are available, it is necessary to make an estimate of
the value of σ that would be obtained were this possible. Similarly, the mean value obtained is less
likely to be the same as that which would be obtained if a very large number of measurements
could be taken, therefore an estimate has to be made of the possible error from the “true” mean.
4.7 Equation (4) gives the standard deviation for the samples actually selected, rather than of the whole
population of possible samples. However, from the results of a single sample of measurements, an
estimate, s(q
j
), can be made of the standard deviation σ of the whole population of possible values of
the measurand from the relation
( ) ( )
2
1
1
1
¿
=
÷
÷
=
n
j
j j
q q
n
q s (5)
4.8
The mean value q will have been derived from a finite number n of samples and therefore its value
will not be the exact mean that would have been obtained if an infinite number of samples could have
been taken. The mean value itself therefore has uncertainty. This uncertainty is referred to as the
experimental standard deviation of the mean. It is obtained from the estimated standard deviation of
the population by the expression:
( )
( )
n
q s
q s
j
=
(6)
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 19 OF 82
4.9 Example: Four measurements were made to determine the repeatability of a measurement system.
The results obtained were 3.42, 3.88, 2.99 and 3.17.
The mean value
4
17 . 3 99 . 2 88 . 3 42 . 3 1
1
+ + +
= =
¿
=
n
j
j
q
n
q = 3.365
The estimated standard deviation ( ) ( )
2
1
1
1
¿
=
÷
÷
=
n
j
j j
q q
n
q s = 0.386
The experimental standard deviation of the mean ( )
( )
193 . 0
4
386 . 0
= = =
n
q s
q s
j
For information on the use of calculators to calculate s(q
j
), see Paragraph P4.
4.10 It may not always be practical or possible to repeat the measurement many times during a test or a
calibration. In these cases a more reliable estimate of the standard deviation of a measurement
system may be obtained from data obtained previously, based on a larger number of readings.
4.11 Whenever possible at least two measurements should be made as part of the procedure; however, it
is acceptable for a single measurement to be made even though it is known that the system has
imperfect repeatability, and to rely on a previous assessment of the repeatability of similar devices.
This procedure must be treated with caution because the reliability of a previous assessment will
depend on the number of devices sampled and how well this sample represents all devices. It is also
recommended that data obtained from prior assessment should be regularly reviewed. Of course,
when only one measurement is made on the device being calibrated a value of s(q
j
) must have been
obtained from prior measurements, and n in Equation (6) is then 1.
In such cases, the estimated standard deviation s(q
j
) is given by
( ) ( )
2
1
1
1
¿
=
÷
÷
=
m
j
j j
q q
m
q s , (7)
4.12
where m is the number of readings considered in the previous evaluation. The standard deviation
of the mean ( )
( )
n
q s
q s
j
= , n being the number of measurements contributing to the reported mean
value.
NOTE
The degrees of freedom under such circumstances are m – 1, where mis the number of measurements in the prior
evaluation. Indeed, this is the reason that a large number of readings in a prior evaluation can give a more reliable estimate
when only a few measurements can be made during the routine procedure. Degrees of freedom are discussed further in
Appendix B.
4.13 The standard uncertainty is then the standard deviation of the mean, i.e. ( ) ( ) q s x u
i
= . (8)
4.14 A previous estimate of standard deviation can only be used if there has been no subsequent
change in the measurement system or procedure that could have an effect on the repeatability. If
an apparently excessive spread in measurement values is found, which is not typical of the
measurement system, the cause should be investigated and resolved before proceeding further.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 20 OF 82
5 TYPE B EVALUATION OF STANDARD UNCERTAINTY
5.1 It is probable that systematic components of uncertainty, i.e. those that account for errors that remain
constant while the measurement is made, will be estimated from Type B evaluations. The most
important of these systematic components, for a reference instrument, will often be the imported
uncertainties associated with its own calibration. However, there can be, and usually are, other
important contributions to systematic errors in measurement that arise in the equipment user's own
laboratory.
5.2 The successful identification and evaluation of these contributions depends on a detailed knowledge
of the measurement process and the experience of the person making the measurements. The need
for the utmost vigilance in preventing mistakes cannot be overemphasised. Common examples are
errors in the corrections applied to values, transcription errors, and faults in software designed to
control or report on a measurement process. The effects of such mistakes cannot readily be included
in the evaluation of uncertainty.
5.3 In evaluating the components of uncertainty it is necessary to consider and include at least the
following possible sources:
(a) The reported calibration uncertainty assigned to reference standards and any drift or instability in their
values or readings.
(b) The calibration of measuring equipment, including ancillaries such as connecting leads etc., and any
drift or instability in their values or readings.
(c) The equipment or item being measured, for example its resolution and any instability during the
measurement. It should be noted that the anticipated longterm performance of the item being
calibrated is not normally included in the uncertainty evaluation for that calibration.
(d)
The operational procedure.
(e)
Variability between different staff carrying out the same type of measurement.
(f)
The effects of environmental conditions on any or all of the above.
5.4 Whenever possible, corrections should be made for errors revealed by calibration or other sources;
the convention is that an error is given a positive sign if the measured value is greater than the
conventional true value. The correction for error involves subtracting the error from the measured
value. On occasions, to simplify the measurement process it may be convenient to treat such an
error, when it is small compared with other uncertainties, as if it were a systematic uncertainty equal
to (±) the uncorrected error magnitude.
5.5 Having identified all the possible systematic components of uncertainty based as far as possible on
experimental data or on theoretical grounds, they should be characterised in terms of standard
uncertainties based on the assessed probability distributions. The probability distribution of an
uncertainty obtained from a Type B evaluation can take a variety of forms but it is generally
acceptable to assign welldefined geometric shapes for which the standard uncertainty can be
obtained from a simple calculation. These distributions and sample calculations are presented in
detail in paragraphs 3.15 to 3.22.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 21 OF 82
6 REPORTING OF RESULTS
6.1 After the expanded uncertainty has been calculated for a coverage probability of 95% the value of
the measurand and expanded uncertainty should be reported as y ± U and accompanied by the
following statement of confidence:
6.2 "The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage
factor k = 2, providing a coverage probability of approximately 95%. The uncertainty evaluation has
been carried out in accordance with UKAS requirements".
6.3 In cases where the procedure of Appendix B has been followed the actual value of the coverage
factor should be substituted for k = 2 and the following statement used:
6.4 "The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage
factor k = XX, which for a tdistribution with v
eff
= YY effective degrees of freedom corresponds to a
coverage probability of approximately 95%. The uncertainty evaluation has been carried out in
accordance with UKAS requirements".
6.5 For the purposes of this document "approximately" is interpreted as meaning sufficiently close that
any difference may be considered insignificant.
6.6 In the special circumstances where a dominant nonGaussian Type B contribution occurs refer to
Appendix C. If uncertainty is being reported as an analytical expression, refer to Appendix L .
6.7 Uncertainties are usually expressed in bilateral terms (±) either in units of the measurand or as
relative values, for example as a percentage (%), parts per million (ppm), 1 in 10
X
, etc. However there
may be situations where the upper and lower uncertainty values are different; for example if cosine
errors are involved. If such differences are small then the most practical approach is to report the
expanded uncertainty as ± the larger of the two. However if there is a significant difference between
the upper and lower values then they should be evaluated and reported separately.
6.8 The number of figures in a reported uncertainty should always reflect practical measurement
capability. In view of the process for estimating uncertainties it is seldom justified to report more than
two significant figures. It is therefore recommended that the expanded uncertainty be rounded to two
significant figures, using the normal rules of rounding. The numerical value of the measurement result
should in the final statement normally be rounded to the least significant figure in the value of the
expanded uncertainty assigned to the measurement result.
6.9
Rounding should always be carried out at the end of the process in order to avoid the effects of
cumulative rounding errors.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 22 OF 82
7 STEP BY STEP PROCEDURE FOR EVALUATION OF MEASUREMENT
UNCERTAINTY
The following is a guide to the use of this code of practice for the treatment of uncertainties. The left
hand column gives the general case while the right hand column indicates how this relates to
example K4 in Appendix K. Although this example relates to a calibration activity, the process for
testing activities is unchanged.
General case
Example K4: Calibration of a weight of
nominal value 10 kg of OIML Class M1
7.1 If possible determine the mathematical
relationship between values of the input
quantities and that of the measurand:
y = f (x
1
, x
2
, … x
N
)
See Appendix D for details.
It will be assumed that the unknown weight, W
X
,
can be obtained from the following relationship:
W
X
= W
S
+ D
S
+ δI
d
+ δC + Ab
7.2 Identify all corrections that have to be applied to
the results of measurements of a quantity
(measurand) for the stated conditions of
measurement.
It is not normal practice to apply corrections
for this class of weight and the comparator
has no measurable linearity error, however,
uncertainties for these contributions have
been determined, therefore:
Drift of standard mass since last calibration:
Correction for air buoyancy:
Linearity correction:
Effect of least significant digit resolution:
0
0
0
0
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 23 OF 82
General case
Example K4: Calibration of a weight of
nominal value 10 kg of OIML Class M1
Source of uncertainty Limit
(mg)
Distribution
W
S
Calibration of std. mass 30 Normal (k = 2)
D
S
Drift of standard mass 30 Rectangular
δC
Comparator linearity 3 Rectangular
δAb
Air buoyancy 10 Rectangular
δI
d
Resolution effects 10 Triangular
7.3 List systematic components of uncertainty
associated with corrections and uncorrected
systematic errors treated as uncertainties.
Seek prior experimental work or theory as a
basis for assigning uncertainties and probability
distributions to the systematic components of
uncertainty.
Calculate the standard uncertainty for each
component of uncertainty, obtained from Type B
evaluations, as in Table 1.
For assumed rectangular distributions:
( )
3
i
a
i
x u =
For assumed triangular distributions:
( )
6
i
a
i
x u =
For assumed normal distributions:
( )
k
U
i
x u =
or consult other documents if the assumed
probability distribution is not covered in this
publication.
Then:
( ) ( )
2
30
1
= =
S
W u x u = 15 mg
( ) ( )
3
30
2
= =
S
D u x u = 17.32 mg
( ) ( )
3
3
3
= = C u x u δ = 1.73 mg
( ) ( )
3
10
4
= = Ab u x u = 5.77 mg
( ) ( )
6
10
5
= =
d
I u x u δ = 4.08 mg
7.4 Use prior knowledge or make trial
measurements and calculations to determine if
there is going to be a random component of
uncertainty that is significant compared with the
effect of the listed systematic components of
uncertainty. Random components of uncertainty
also have to be considered as input quantities.
From previous knowledge of the measurement
system it is known that there is a significant
random component of uncertainty.
7.5 If a random component of uncertainty is
significant make repeated measurements to
obtain the mean from Equation (3):
¿
=
=
n
j
j
q
n
q
1
1
Three measurements were made of the
difference between the unknown weight and the
standard weight, from which the mean
difference was calculated:
3
020 . 0 025 . 0 015 . 0 + +
=
S
W = 0.020 g
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 24 OF 82
General case
Example K4: Calibration of a weight of
nominal value 10 kg of OIML Class M1
7.6 Either calculate the standard deviation of the
mean value from Equations (5) and (6):
( ) ( )
2
1
1
1
¿
=
÷
÷
=
n
j
j j
q q
n
q s
( )
( )
n
q s
q s
j
=
or refer to the results of previous repeatability
evaluations for an estimate of s(q
j
) based on a
larger number of readings, using Equation (7):
( ) ( )
2
1
1
1
¿
=
÷
÷
=
m
j
j j
q q
m
q s
( )
( )
n
q s
q s
j
=
where m is the number of readings used in the
prior evaluation and n is the number of readings
that contribute to the mean value.
A previous Type A evaluation had been made to
determine the repeatability of the comparison
using the same type of 10 kg weights. The
standard deviation was determined from 10
measurements using the conventional
bracketing technique and was calculated, using
Equation (5), to be 8.7 mg.
Since the number of determinations taken when
calibrating the unknown weight was 3 this is the
value of n that is used to calculate the standard
deviation of the mean using Equation (6):
( )
( )
3
7 . 8
= =
n
W s
W s
X
= 5.0 mg
7.7 Even when a random component of uncertainty
is not significant, where possible check the
instrument indication at least once to minimise
the possibility of unexpected errors.
7.8 Derive the standard uncertainty for the above
Type A evaluation from Equation (8):
( ) ( ) q s x u
i
=
This is then the standard uncertainty for the
Type A evaluation:
( ) ( ) ( )
R R
W s W u x u = =
6
= 5.0 mg
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 25 OF 82
General case
Example K4: Calibration of a weight of
nominal value 10 kg of OIML Class M1
7.9
Calculate the combined standard uncertainty for
uncorrelated input quantities using Equation (1) if
absolute values are used:
( ) ( ) ( )
¿ ¿
= =
÷ =
N
i
i
N
i
i i c
y u x u c y u
1
2
1
2
2
, where c
i
is the
partial derivative
i
x f c c , or a known sensitivity
coefficient.
Alternatively use Equation (11) if the standard
uncertainties are relative values:
( ) ( )
¿
= (
(
¸
(
¸
=
N
i i
i i
c
x
x u p
y
y u
1
2
, where p
i
are known
positive or negative exponents in the functional
relationship.
The units of all standard uncertainties are in
terms of those of the measurand, i.e. milligrams,
and the functional relationship between the
input quantities and the measurand is a linear
summation; therefore all the sensitivity
coefficients are unity (c
i
=1).
7.10 If correlation is suspected use the guidance in
paragraph D3 or consult other referenced
documents.
None of the input quantities is considered to be
correlated to any significant extent; therefore
Equation (1) can be used to calculate the
combined standard uncertainty:
( )
2 2 2 2 2 2
0 . 5 77 . 5 73 . 1 08 . 4 32 . 17 15 + + + + + =
X
W u
= 24.55 mg.
7.11 Either calculate an expanded uncertainty from
Equation (2):
U = k·u
c
(y)
or, if there is a significant random contribution
evaluated from a small number of readings, use
Appendix Bto calculate a value for k
p
and use
this value to calculate the expanded uncertainty.
U= 2 x 24.55 mg = 49.10 mg
It was not necessary to use Appendix B to
determine a value for k
p
. In fact the effective
degrees of freedom of u(W
X
) are greater than
5000 which gives a value for k
95
= 2.00.
7.12 Report the result and the expanded uncertainty
in accordance with Section 6.
The measured value of the 10 kg weight is
10 000.025 g ± 0.049 g.
The reported expanded uncertainty is based on a standard
uncertainty multiplied by a coverage factor k = 2, providing a
coverage probability of approximately 95%. The uncertainty
evaluation has been carried out in accordance with UKAS
requirements.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 26 OF 82
APPENDIX A
BEST MEASUREMENT CAPABILITY
A1 Best measurement capability (BMC) is a term normally used to describe the uncertainty that appears
in an accredited calibration laboratory's schedule of accreditation and is the uncertainty for which the
laboratory has been accredited using the procedure that was the subject of assessment. The best
measurement capability should be calculated according to the procedures given in this document and
should normally be quoted as an expanded uncertainty at a coverage probability of 95%, which
usually requires the use of a coverage factor of k = 2.
A2 An accredited laboratory is not permitted to report an uncertainty smaller than their accredited best
measurement capability but may report an equal or larger uncertainty. Since the magnitude of the
uncertainty reported on a certificate of calibration will often depend on properties of the device being
calibrated any definition of best measurement capability should not include uncertainties that are
dependent on this device. However, no device is perfect and so the concept of “nearly ideal” is used
in association with the evaluation of a BMC. It may also be the case that an accredited laboratory can
achieve a particular uncertainty if conditions are optimum but cannot achieve this uncertainty
routinely.
A3 In order to promote harmony between accredited laboratories and between accreditation bodies the
EAhas adopted the following definition of best measurement capability:
The smallest uncertainty of measurement a laboratory can achieve within its scope of accreditation,
when performing more or less routine calibrations of nearly ideal measurement standards intended to
define, realize, conserve or reproduce a unit of that quantity or one or more of its values, or when
performing more or less routine calibrations of nearly ideal measuring instruments designed for the
measurement of that quantity.
In other words "best measurement capability" is the smallest uncertainty a laboratory can achieve
when performing more or less routine calibrations on a nearly ideal device being calibrated.
A4 Anearly ideal device is one that is available but does not necessarily represent the majority of
devices that the laboratory may be asked to calibrate. The properties of these devices that are
considered to be nearly ideal will depend on the field of calibration but may include an instrument with
very low random fluctuations, negligible temperature coefficient, very low voltage reflection coefficient
etc. The uncertainty budget that is intended to demonstrate the best measurement uncertainty should
still include contributions from the properties of the device being calibrated that are considered to be
nearly ideal but the value of the uncertainty may be entered as zero or a negligible value, if this is the
case. Where necessary the laboratory's schedule of accreditation will include a remark that describes
the conditions under which the best measurement capability can be achieved.
A5 By "more or less routine calibrations" it is meant that the laboratory shall be able to achieve the stated
capability in the normal work that it performs under its accreditation and, by implication, using the
procedures, equipment and facilities that were the subject of the assessment. Where a smaller
uncertainty can be achieved by, for example, taking a large number of readings this should be
considered when arriving at the budget for the best measurement capability and would therefore be
within the "more or less routine" conditions.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 27 OF 82
A6 It is sometimes the case that a laboratory may wish to be accredited for a measurement uncertainty
that is larger than it can actually achieve. If the principles of this document are followed when
constructing the uncertainty budget the resulting expanded uncertainty should be a realistic
representation of the laboratory’s measurement capability. If this is smaller than the uncertainty the
laboratory wishes to be accredited for and report on their certificates of calibration, the implication is
that the laboratory is uncomfortable in some way about the magnitude of the expanded uncertainty. If
this is the case then the contributions to the uncertainty budget should be reviewed and consideration
given to making more conservative allowances as necessary.
A7 It can be the case that some calibration laboratories offer a best measurement capability with very
small uncertainties but these are not routinely offered for everyday calibrations. This is because
they will maintain their own reference standards, upon which the BMC is based, but use subsidiary
equipment – often automated – for routine work. The contract review arrangements between the
laboratory and its customer should define the level of service being offered.
A8 In some cases the best measurement capability quoted in a laboratory's schedule has to cover a two
(or more) dimensional range of measured values, such as different levels and frequencies, and it may
not be practical to give the actual uncertainty for all possible values of the quantities. In these cases
the best measurement capability may be given as a range of uncertainties appropriate to the upper
and lower values of the uncertainty that has been calculated for the range of the quantity, or may be
described as an expression. Guidance about the expression of uncertainty over a range of values is
presented in Appendix L.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 28 OF 82
APPENDIX B
DERIVINGACOVERAGE FACTOR FOR UNRELIABLE INPUT QUANTITIES
B1 In the majority of measurement situations it will be possible to evaluate Type B uncertainties with high
reliability. Furthermore, if the procedure followed for making the measurements is well established
and if the Type Aevaluations are obtained from a sufficient number of observations then the use of a
coverage factor of k = 2 will mean that the expanded uncertainty, U, will provide an interval with a
coverage probability close to 95%. This is because the distribution tends to normality as the number
of observations increases and k = 2 corresponds to 95% confidence for a normal distribution.
B2 However, in some cases it may not be practical to base the Type Aevaluation on a large number of
readings, which could result in the coverage probability being significantly less than 95% if a
coverage factor of k = 2 is used. In these situations the value of k, or more correctly k
p
, where p is the
confidence probability, should be based on a tdistribution rather than a normal distribution. This value
of k
p
will give an expanded uncertainty, U
p
, that maintains the coverage probability at approximately
the required level p.
y – k y y + k
Figure 6
In Figure 6, the solid line indicates the normal distribution. A specified
proportion p of the values under the curve are encompassed between y – k
and y + k. An example of the tdistribution is superimposed, using dashed
lines. For the tdistribution, a greater proportion of the values lie outside the
region y – k to y + k, and a smaller proportion lie inside this region. An
increased value of k is therefore required to restore the original coverage
probability. This new coverage factor, k
p
, is obtained by evaluating the
effective degrees of freedom of u
c
(y) and obtaining the corresponding value of
t
p
, and hence k
p
, from the tdistribution table.
B3 In order to obtain a value for k
p
it is necessary to obtain an estimate of the effective degrees of
freedom, v
eff
, of the combined standard uncertainty u
c
(y). The GUM recommends that the Welch
Satterthwaite equation is used to calculate a value for v
eff
based on the degrees of freedom, v
i
, of the
individual standard uncertainties u
i
(y); therefore
( )
( )
¿
=
=
N
i
i
i
c
eff
y u
y u
1
4
4
ν
ν
(9)
B4 The degrees of freedom, ν
i
, for contributions obtained from Type Aevaluations are n  1, where n is
the number of readings used to evaluate s(q
j
).
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 29 OF 82
B5 It is often possible to take the degrees of freedom, ν
i
, of Type B uncertainty contributions as infinite,
that is, their value is known with a very high degree of reliability. If this is the case, and there is only
one contribution obtained from a Type Aevaluation, then the process using the WelchSatterthwaite
formula simplifies, as all the terms relating to the type B uncertainties become zero. This is illustrated
in the example in paragraph B10.
B6 However, a Type B contribution may have come from a calibration certificate as an expanded
uncertainty based on a tdistribution rather than a normal distribution, as described in this Appendix.
This is an example of a Type B uncertainty that does not have infinite degrees of freedom. For this
eventuality the degrees of freedom will be as quoted on the calibration certificate, or it can be
obtained from the tdistribution table for the appropriate value of k
95
.
B7 Having obtained a value for v
eff
, the tdistribution table is used to find a value of k
p
. This table,
reproduced below, gives values for k
p
, for various levels of confidence p. Unless otherwise specified,
the values corresponding to p = 95.45% should be used.
Values of t
p
(ν) from the tdistribution for degrees of freedom ν that define an interval
that encompasses specified fractions p of the corresponding distribution
Degrees of
freedom
ν p = 68.27% p = 90% p = 95% p = 95.45% p = 99% p = 99.73%
1 1.84 6.31 12.71 13.97 63.66 235.80
2 1.32 2.92 4.30 4.53 9.92 19.21
3 1.20 2.35 3.18 3.31 5.84 9.22
4 1.14 2.13 2.78 2.87 4.60 6.62
5 1.11 2.02 2.57 2.65 4.03 5.51
6 1.09 1.94 2.45 2.52 3.71 4.90
7 1.08 1.89 2.36 2.43 3.50 4.53
8 1.07 1.86 2.31 2.37 3.36 4.28
9 1.06 1.83 2.26 2.32 3.25 4.09
10 1.05 1.81 2.23 2.28 3.17 3.96
11 1.05 1.80 2.20 2.25 3.11 3.85
12 1.04 1.78 2.18 2.23 3.05 3.76
13 1.04 1.77 2.16 2.21 3.01 3.69
14 1.04 1.76 2.14 2.20 2.98 3.64
15 1.03 1.75 2.13 2.18 2.95 3.59
16 1.03 1.75 2.12 2.17 2.92 3.54
17 1.03 1.74 2.11 2.16 2.90 3.51
18 1.03 1.73 2.10 2.15 2.88 3.48
19 1.03 1.73 2.09 2.14 2.86 3.45
20 1.03 1.72 2.09 2.13 2.85 3.42
25 1.02 1.71 2.06 2.11 2.79 3.33
30 1.01 1.70 2.04 2.09 2.75 3.27
35 1.01 1.70 2.03 2.07 2.72 3.23
40 1.01 1.68 2.02 2.06 2.70 3.20
45 1.01 1.68 2.01 2.06 2.69 3.18
50 1.01 1.68 2.01 2.05 2.68 3.16
100 1.005 1.660 1.984 2.025 2.626 3.077
∞ 1.000 1.645 1.960 2.000 2.576 3.000
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 30 OF 82
B8 Normally v
eff
will not be an integer and it will be necessary to interpolate between the values given in
the table. Linear interpolation will suffice for v
eff
>3; higherorder interpolation should be used
otherwise. Alternatively, the next lower value of v
eff
may be used.
B9 The value of t
p
(ν) obtained from the table is the coverage factor k
p
that is required to calculate the
expanded uncertainty, U
p
, from U
p
= k
p
u
c
(y). Unless otherwise specified, the coverage probability p
will usually be 95%.
B10 Example
B10.1
In a measurement system a Type Aevaluation, based on 4 observations, gave a value of u
i
(y) of 3.5
units using Equations (4) and (5). There were 5 other contributions all based on Type B evaluations
for each of which infinite degrees of freedom had been assumed. The combined standard uncertainty,
u
c
(y), had a value of 5.7 units. Then using the WelchSatterthwaite formula (paragraph B3, Equation
9):
1 . 21 3
5 . 3
7 . 5
0 0 0 0 0
1 4
5 . 3
7 . 5
4
4 4
4
= =
+ + + + +
÷
= x
eff
u
B10.2 The value of v
eff
given in the tdistribution table, for a coverage probability p of 95.45%, immediately
lower than 21.1 is 20. This gives a value for k
p
of 2.13 and this is the coverage factor that should be
used to calculate the expanded uncertainty. The expanded uncertainty is 5.7 x 2.13 = 12.14 units.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 31 OF 82
APPENDIX C
DOMINANT NONGAUSSIAN TYPE B UNCERTAINTY
C1.1 In some measurement processes there can be one component of uncertainty derived from a Type B
evaluation that is dominant in magnitude compared with the other components. When the dominant
component is characterised by limits within which there is a high probability of occurrence, a
calculated expanded uncertainty, U, using the coverage factor of k = 2, may be greater than the
arithmetic sum of the semirange of all the individual limiting values. As it is reasonable to assume
that the arithmetic sum of these contributions would be for a coverage probability approaching 100%,
there is a degree of pessimism in following the normal recommended procedure for combination of
uncertainties.
C1.2 Consequently special consideration needs to be given to the situation in which the calculated
expanded uncertainty fails to meet the criterion U s arithmetic sum of the limit values of all
contributions. For practical purposes, the “limit” value of a normal distribution can be taken as three
times its standard deviation.
C1.3 In many cases this criterion will be met, but it can be the case that a single rectangular distribution
may dominate over other contributions, Acommonly encountered example is the resolution of a
digital indicating instrument. This will have a standard uncertainty of
3
i
a
. If this is large compared to
other contributions it is less likely that the criterion above will be met. When it is not met then the
dominant contribution, a
d
, should be extracted and a new value of the expanded uncertainty should
be reported as U = U’ + a
d
, where U’ is the expanded uncertainty for the remaining components,
treated in the normal statistical manner.
C1.4 The use of a twopart expression such as this means that when both components are imported into
a subsequent uncertainty budget it is likely that a
d
will no longer be a dominant component and a
normal distribution can be assumed for the subsequent combined standard uncertainty.
C1.5 Under these circumstances, the statement of coverage probability associated with the expanded
uncertainty will have to be modified; an example is given below:
The uncertainty is stated in two parts, the second of which is dominant and is due to the uncertainty due to the
resolution of the instrument being calibrated, for which a rectangular probability distribution has been assumed.
C1.6 The same situation may be encountered with other distributions associated with Type B uncertainties.
An example is the Ushaped distribution associated with mismatch uncertainty in RF and microwave
systems. Similar reasoning applies here, and the suggested coverage probability statement can be
modified accordingly.
C1.7 There may be, however, situations where a single value of uncertainty is required even if there is a
Type B uncertainty that causes the distribution to be nonnormal. This will involve evaluation of a
coverage factor for a stated coverage probability for the convolved distributions.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 32 OF 82
C1.8 If a rectangular distribution and a normal distribution are convolved, the coverage factor k for a
coverage probability of 95.45% may be obtained from the following table:
( )
( )
rect
i
normal
i
y u
y u
45 . 95
k
( )
( )
rect
i
normal
i
y u
y u
45 . 95
k
( )
( )
rect
i
normal
i
y u
y u
45 . 95
k
0.00 1.65 0.50 1.84 0.95 1.95
0.10 1.66 0.55 1.85 1.00 1.95
0.15 1.68 0.60 1.87 1.10 1.96
0.20 1.70 0.65 1.89 1.20 1.97
0.25 1.72 0.70 1.90 1.40 1.98
0.30 1.75 0.75 1.91 1.80 1.99
0.35 1.77 0.80 1.92 2.00 1.99
0.40 1.79 0.85 1.93 2.50 2.00
0.45 1.82 0.90 1.94 ∞ 2.00
C1.9 Example
A digital voltmeter is calibrated with an applied voltage of 1.0000 V; the resulting reading is 1.001 V.
The expanded uncertainty of the applied voltage is ± 0.0002 V (k=2). The only other uncertainty of
significance is due to the rounding of the indicator display. The indicator can display readings in steps
of 0.001 V; therefore there will be a possible rounding error of ± 0.0005 V. A rectangular probability
distribution is assumed.
The uncertainty budget will therefore be as follows:
Symbol Source of uncertainty Value
± V
Probability distribution Divisor c
i
u
i
(V)
V
ν
i
or
ν
eff
V
s
Uncertainty of applied voltage 0.0002 Normal 2 1 0.000100 ∞
δI
d
Digital rounding of indicator 0.0005 Rectangular √3 1 0.000289 ∞
u
c
(V) Combined standard uncertainty
Convolved
346 . 0
000289 . 0
000100 . 0
=
0.000306 ∞
U Expanded uncertainty
Convolved
k = 1.77
0.000541 ∞
Reported result
For an applied voltage of 1.000 V the voltmeter reading was 1.001 V ± 0.00054V.
The reported expanded uncertainty is based on a convolution of a dominant rectangular uncertainty with other,
smaller, uncertainties. The resulting standard uncertainty has been multiplied by a coverage factor k = 1.77, which,
for this particular convolution, corresponds to a coverage probability of approximately 95%. The uncertainty
evaluation has been carried out in accordance with UKAS requirements.
NOTE
Asimple test to determine whether a rectangular uncertainty is a dominant component is to check whether its standard uncertainty
is more than 1.4 times the combined standard uncertainty for the remaining components. If it is not, then
( )
( )
71 . 0 >
rect
y
i
u
normal
y
i
u
and
the coverage factor k will be within 5% of the usual value of 2.00.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 33 OF 82
C1.10 If a Ushaped distribution and a normal distribution are convolved, the coverage factor k for a
coverage probability of 95.45% may be obtained from the following table:
( )
( )
shaped U
i
normal
i
y u
y u
÷
45 . 95
k
( )
( )
shaped U
i
normal
i
y u
y u
÷
45 . 95
k
( )
( )
shaped U
i
normal
i
y u
y u
÷
45 . 95
k
0.00 1.41 0.50 1.77 0.95 1.93
0.10 1.47 0.55 1.80 1.00 1.93
0.15 1.51 0.60 1.82 1.10 1.95
0.20 1.55 0.65 1.84 1.20 1.96
0.25 1.60 0.70 1.86 1.40 1.97
0.30 1.64 0.75 1.88 1.80 1.99
0.35 1.67 0.80 1.89 2.00 1.99
0.40 1.71 0.85 1.90 2.50 2.00
0.45 1.74 0.90 1.92 ∞ 2.00
C1.11 If a Ushaped distribution and a rectangular distribution are convolved, the coverage factor k for a
coverage probability of 95.45% may be obtained from the following table:
( )
( )
shaped U i
rect i
y u
y u
÷
45 . 95
k
( )
( )
shaped U i
rect i
y u
y u
÷
45 . 95
k
( )
( )
shaped U i
rect i
y u
y u
÷
45 . 95
k
0.00 1.41 0.45 1.75 3.0 1.80
0.10 1.48 0.50 1.78 4.0 1.75
0.15 1.53 0.60 1.82 5.0 1.72
0.20 1.57 0.70 1.86 6.0 1.70
0.25 1.62 0.80 1.88 7.5 1.68
0.30 1.66 0.90 1.89 10 1.66
0.35 1.69 1.0 1.90 20 1.65
0.40 1.73 2.0 1.86 ∞ 1.65
C1.12 If two distributions of identical form, either rectangular or Ushaped, are convolved, the coverage
factor k for a coverage probability of 95.45% may be obtained from the following table:
Ratio
( )
( )
larger
i
smaller
i
y u
y u
k for stated ratio
2 Rectangular Distributions
k for stated ratio
2 Ushaped Distributions
0.00 1.65 1.41
0.05 1.65 1.44
0.10 1.66 1.49
0.15 1.69 1.53
0.20 1.71 1.58
0.25 1.74 1.62
0.30 1.77 1.66
0.35 1.79 1.69
0.40 1.82 1.72
0.45 1.84 1.75
0.50 1.86 1.77
0.60 1.89 1.81
0.70 1.91 1.83
0.80 1.92 1.85
0.90 1.93 1.86
1.00 1.93 1.86
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 34 OF 82
APPENDIX D
DERIVATION OF THE MATHEMATICAL MODEL
D1 Measurement process
D1.1 In general a measurement process can be regarded as having estimated input quantities X
i
whose
values, given the symbol x, contribute to the estimated value y of the measurand or output quantity Y.
Where, as in most cases, there are several input quantities, the standard uncertainty associated with
the estimated value of each one is represented by u(x
i
). Standard uncertainty and its evaluation are
discussed in Sections 4 and 5.
D1.2 The measurement process can usually be modelled by a functional relationship between the values
of the estimated input quantities and that of the output estimate in the form
y = f (x
1
, x
2
, … x
N
)
For example, if electrical resistance R is measured in terms of voltage V and current I then the
relationship is R = f(V,I) = V/I. The mathematical model of the measurement process is used to
identify the input quantities that need to be considered in the uncertainty budget and their relationship
to the total uncertainty for the measurement.
D1.3 Example
The following example of how a mathematical model can be derived relates to example K1 in
Appendix K.
In this example two resistors, R
S
and R
X
, are compared by connecting them in series and passing a
constant current through them. The voltage across each is measured. As the same current passes
through both resistors the ratio of the two voltages V
S
and V
X
will be the same as the ratio of the two
resistance values, i.e.
S
X
S
X
V
V
R
R
=
If the resistance of R
S
is known the value of R
X
can be determined by rearranging the equation as
follows:
S
X
S X
V
V
R R · =
It is known that there will be various uncertainties associated with the measurement. In the
expression above, the calibration uncertainty of R
S
means that the same relative uncertainty will exist
in the value of R
X
. There are, however, further uncertainties to be considered. For example, the value
of R
S
will change with time, therefore a contribution δR
D
has to be included – the model now
becomes:
( )
S
X
D S X
V
V
R R R · + = δ
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 35 OF 82
Another consideration is the effect of temperature. The temperature coefficient of resistors of this type
is usually described by a parabolic curve. Taking the partial derivatives of the expression describing
this curve is not a straightforward process; therefore the practical approach of estimated the
temperature coefficient at the nominal temperature was taken, using a linear approximation. This
value for the temperature coefficient, R
TC
, is multiplied by the value assigned for temperature
variations, Δt, and the product is inserted in the model:
( ) ( )
S
X
TC D S X
V
V
t R R R R · · + + = Δ δ
Another influence is the ratio V
X
/V
S
. In this case, the same voltmeter is used to measure both V
S
and
V
X
. Any systematic error, or offset, in the voltmeter reading will cancel because it is the same for both
voltages  the ratio will be unchanged. All that have to be considered are secondary influences, such
as resolution and linearity, and it can be seen from examination of the uncertainty budget for example
K1 that these influences have been considered separately. This is a good example of negative
correlation.
The repeatability of the process also has to be included. As the parameters directly observed during
the calibration are the voltages V
X
and V
S
, it is convenient to assess the repeatability in terms of
observed changes in the ratio V
X
/V
S
. This is carried out in accordance with Equations (4) and (5),
which yield the experimental standard deviation of the mean, ( ) V s . The model then becomes:
( ) ( ) ( )
S
X
TC D S X
V
V
V s t R R R R · + · + + = Δ δ
This is how the model in Example K1 was derived.
D2 Model descriptions
D2.1 If the functional relationship is an addition or subtraction of the input quantities, for example
W
X
= f (W
S
,D
S
,δI
d
,δC, Ab) = W
S
+ D
S
+ δI
d
+ δC + Ab,
then all the input quantities will be directly related to the output quantity and the partial derivatives will
all be unity.
D2.2 In paragraph 3.35 it is stated that the combined standard uncertainty is calculated using the
expression
( ) ( ) ( )
¿ ¿
= =
÷ =
N
i
i
N
i
i i c
y u x u c y u
1
2
1
2
2
(10)
D2.3 If the functional relationship is a product or quotient, i.e. the output quantity is obtained from only the
multiplication or division of the input quantities, this can be transformed to a linear addition by the use
of relative values, e.g. those expressed in % terms or in parts per million. The general form is
N
p
N
p p
x x cx y · · · · =
2 1
2 1
where the exponents p
i
are known positive or negative numbers. The standard
uncertainty will then be given by
( ) ( )
¿
= (
(
¸
(
¸
=
N
i i
i i c
x
x u p
y
y u
1
2
(11)
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 36 OF 82
D2.4 Some examples of the use of relative uncertainties are
P = f(V, I) = V·I , so
( ) ( ) ( )
2 2
(
¸
(
¸
+
(
¸
(
¸
=
I
I u
V
V u
P
P u
P = f(V, R) = V
2
/R, so
( ) ( ) ( )
2 2
2
(
¸
(
¸
+
(
¸
(
¸
=
R
R u
V
V u
P
P u
V = f(P, Z) = (P·Z)
½
, so
( ) ( ) ( )
2 2
2 2
(
¸
(
¸
+
(
¸
(
¸
=
Z
Z u
P
P u
V
V u
D2.5 The use of relative uncertainties can often simplify the calculations and is particularly helpful when the
input quantities and the uncertainties are already given in relative terms. However, sensitivity
coefficients may still be required to account for known relationships, such as a temperature
coefficient. Relative uncertainties should not be used when the functional relationship is already an
addition or subtraction.
D3 Correlated input quantities
D3.1 The expressions given for the standard uncertainty of the output estimate, Equations (10) and (11),
will only apply when there is no correlation between any of the input estimates, that is, the input
quantities are independent of each other. It may be the case that some input quantities are affected
by the same influence quantity, e.g. temperature, or by the errors in a particular instrument that is
used for separate measurements in the same process. In such cases the input quantities are not
independent of each other and the equation for obtaining the standard uncertainty of the output
estimate must be modified.
D3.2 The effects of correlated input quantities may serve to reduce the combined standard uncertainty,
such as when an instrument is used as a comparator between a standard and an unknown, and this
is referred to as negative correlation. In other cases measurement errors will always combine in one
direction and this has to be accounted for by an increase in the combined standard uncertainty. This
is referred to as positive correlation. Knowledge concerning the possibility of correlation can often be
obtained from the functional relationship between the input quantities and the output quantity but it
may also be necessary to investigate the effects of correlation by making a planned series of
measurements.
D3.3 If positive correlation between input quantities is suspected but the degree of correlation cannot
easily be determined then the most straightforward solution is to add arithmetically the standard
uncertainties for these quantities to give a new standard uncertainty that is then dealt with in the usual
manner in Equation (1) or (11). The Guide should be consulted for a more detailed approach to
dealing with correlation based on the calculation of correlation coefficients.
D3.4 An example of the treatment of correlated contributions is shown in paragraph K6.4.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 37 OF 82
APPENDIX E
SOME SOURCES OF ERROR AND UNCERTAINTY IN ELECTRICAL CALIBRATIONS
The following is a description of the more common sources of systematic error and uncertainty (after
correction) in electrical calibration work, with brief comments about their nature. Further, more detailed,
advice is given in specialised technical publications and manufacturers’ application notes, as well as other
sources.
E1 Imported uncertainty
E1.1 The uncertainties assigned to the values on a calibration certificate for the calibration of an instrument,
whether measuring equipment or a reference standard, are all contributors to the uncertainty budget.
E2 Secular stability
E2.1 The performance of all instruments, and the values of reference standards, must be expected to change
to some extent with the passage of time. Passive devices such as standard resistors or highgrade RF
and microwave attenuators may be expected to drift slowly with time. An estimate of such a drift has to
be assessed on the basis of values obtained from previous calibrations. It cannot be assumed that a drift
will be linear. Data can be assimilated readily if displayed in a graphical form. A curve fitting procedure
that gives a progressively greater weight to each of the more recent calibrations can be used to allow the
most probable value at the time of use to be assessed. The degree of complexity in curve fitting is a
matter of judgement; in some cases drawing a smooth curve through the chosen data points by hand
can be quite satisfactory. Whenever a new calibration is obtained the drift characteristic will need re
assessment. The corrections that are applied for drift are subject to uncertainty based on the scatter of
data points about the drift characteristic. The magnitude of the drift and the random instability of an
instrument, and the accuracy required will determine the calibration interval.
E2.2 With complex electronic equipment it is not always possible to follow this procedure as changes in
performance can be expected to be more random in nature over relatively long periods. Checks against
passive standards can establish whether compliance to specification is being maintained or whether a
calibration with subsequent equipment adjustment is needed. The manufacturer’s specification can be a
good starting point for assigning the uncertainty due to instrument drift, but should be confirmed by
analysis of quality control and calibration data.
E3 Environmental conditions
E3.1 The laboratory measurement environment can be one of the most important considerations when
performing electrical calibrations. Ambient temperature is often the most important influence and
information on the temperature coefficient of, for example, resistance standards has to be sought or
determined. Variations in relative humidity can also affect the values of unsealed components. The
influence of barometric pressure on certain electrical measurement standards can also be significant. At
RF and microwave frequencies, ambient temperature can affect the performance of, for example,
attenuators, impedance standards that depend on mechanical dimensions for their values and other
precision components. Devices that incorporate thermal sensing, such as power sensors, can be affected
by rapid temperature changes that can be introduced by handling or exposure to sunlight or other sources
of heat.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 38 OF 82
E3.2 It is also necessary to be aware of the possible effects of electrical operating conditions, such as power
dissipation, harmonic distortion, or level of applied voltage being different when a device is in use from
when it was calibrated. Resistance standards, resistive voltage dividers and attenuators at any frequency
are examples of devices being affected by selfheating and/or applied voltage. It should also be ensured
that all equipment is operating within the manufacturer's stated range of supply voltages.
E3.3 The effects of harmonics and noise on ac calibration signals may have an influence on the apparent value
of these signals. Similarly, the effects of any commonmode signals present in a measurement system may
have to be accounted for.
E4 Interpolation of calibration data
E4.1 When an instrument with a broad range of measurement capabilities is calibrated there are practical and
economic factors that limit the number of calibration points. Consequently the value of the quantity to be
measured and/or its frequency may be different from any of the calibration points. When the value of the
quantity lies between two calibration values, consideration needs to be given to systematic errors that arise
from, for example, scale nonlinearity.
E4.2 If the measurement frequency falls between two calibration frequencies, it will also be necessary to assess
the additional uncertainty due to interpolation that this can introduce. One can only proceed with
confidence if:
(a) a theory of instrument operation is known from which one can predict a frequency characteristic, or there is
additional frequency calibration data from other models of the same instrument,
and wherever reasonable,
(b) the performance of the actual instrument being used has been explored with a swept frequency
measurement system to verify the absence of resonance effects or aberrations due to manufacturing or
other performance limitations.
E5 Resolution
E5.1 The limit to the ability of an instrument to indicate small changes in the quantity being measured, referred
to as resolution or “digital rounding error”, is treated as a systematic component of uncertainty.
E5.2 Many instruments with a digital display use an analogueto digital converter (ADC) to convert the
analogue signal under investigation into a form that can be displayed in terms of numeric digits. The last
displayed digit will be a rounded representation of the underlying analogue signal. The error introduced
by this process will be from  0.5 digit (else the last digit would be one lower) to + 0.5 digit (else the last
digit would be one higher). A quantisation error of ± 0.5 digit is therefore present. As there is no way of
knowing where within this range the underlying value is, the resulting error is assumed to be zero with
limits of ± 0.5 digit.
E5.3 This “digital rounding error” of ± 0.5 digit may not apply in all instances and an understanding of
instrument operation is needed if the assigned uncertainty is to be realistic. For example, a directgating
frequency counter has a digital rounding error of ± 1 digit, due to the random relationship between the
signal being measured and the internal clock. Some instruments may also display hysteresis that,
although not necessarily a property of the display itself, may result in further uncertainties amounting to
several digits.
E5.4 In an analogue instrument the effect of resolution is determined by the practical ability to read the position
of a pointer on a scale. In either case, the last digit actually recorded will always be subject to an
uncertainty of at least ± 0.5 digit. The presence of electrical noise causing fluctuations in instrument
readings will commonly determine the usable resolution, however it is possible to make a good estimate of
the mean position of a fluctuating pointer by eye.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 39 OF 82
E6 Apparatus layout
E6.1 The physical layout of one item of equipment with respect to another and the relationship of these items to
the earth plane can be important in some measurements. Thus a different arrangement between
calibration and subsequent use of an instrument may be the source of systematic errors. The main effects
are leakage currents to earth, interference loop currents, and electromagnetic leakage fields. In inductance
measurements it is necessary to define connecting lead configuration and be conscious of the possible
effects of an earth plane or adjacent ferromagnetic material. The effect of mutual heating between
apparatus may also need to be considered.
E7 Thermoelectric voltages
E7.1 If an electrical conductor passes through a temperature gradient then a potential difference will be
generated across that gradient. This is known as the Seebeck effect and these unwanted, parasitic
voltages can cause errors in some measurement systems – in particular, where small dc voltages are
being measured.
E7.2 They can be minimised by design of connections that are thermally symmetrical, so that the Seebeck
voltage in one lead is cancelled by an identical and opposite voltage in the other. In some situations, e.g.
ac/dc transfer measurements, the polarity of the dc supply is reversed and an arithmetic mean is taken of
two sets of dc measurements.
E7.3 Generally an allowance has to be made as a Type B component of uncertainty for the presence of thermal
emfs.
E8 Loading and cable impedance
E8.1 The finite input impedance of voltmeters, oscilloscopes and other voltage sensing instruments may so load
the circuit to which they are connected as to cause significant systematic errors. Corrections may be
possible if impedances are known. In particular, it should be noted that some multifunction calibrators can
exhibit a slightly inductive output impedance. This means that when a capacitive load is applied, the
resulting resonance may cause the output voltage to increase with respect to its opencircuit value.
E8.2 The impedance and finite electrical length of connecting leads or cables may also result in systematic
errors in voltage measurements at any frequency. The use of fourterminal connections minimises such
errors in some dc and ac measurements.
E8.3 For capacitance measurements, the inductive properties of the connecting leads may be important,
particularly at higher values of capacitance and/or frequency. Similarly, for inductance measurements the
capacitance between connecting leads may be important.
E9 RF mismatch errors and uncertainty
E9.1 At RF and microwave frequencies the mismatch of components to the characteristic impedance of the
measurement system transmission line can be one of the most important sources of error and of the
systematic component of uncertainty in power and attenuation measurements. This is because the phases
of voltage reflection coefficients are not usually known and hence corrections cannot be applied.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 40 OF 82
E9.2 In a power measurement system, the power, P
O
, that would be absorbed in a load equal to the
characteristic impedance of the transmission line has been shown
[3]
to be related to the actual power, P
L
,
absorbed in a wattmeter terminating the line by the equation

.

\

I I + I I ÷
I ÷
=
2 2
2
0
cos 2 1
1
L G L G
L
L
P
P 
E(1)
where φ is the relative phase of the generator and load voltage reflection coefficients Γ
G
and Γ
L
. When
Γ
G
and Γ
L
are small, this becomes
( )  cos 2 1
1
2
0 L G
L
L
P
P I I ÷
I ÷
=
E(2)
E9.3 When φ is unknown, this expression for absorbed power can have limits:
P
0
(limits) = ( )
L G
L
L
P
I I ±
I ÷
2 1
1
2
E(3)
E9.4
The calculable mismatch error is 1  Γ
L

2
and is accounted for in the calibration factor, while the limits of
mismatch uncertainty are ± 2 Γ
L
 Γ
G
. Because a cosine function characterises the probability
distribution for the uncertainty, Harris and Warner
[3]
show that the distribution is Ushaped with a
standard deviation given by
u(mismatch) =
L G
L G
I I =
I I
414 . 1
2
2
E(4)
E9.5 When a measurement is made of the attenuation of a twoport component inserted between a generator
and load that are not perfectly matched to the transmission line, Harris and Warner
[3]
have shown that the
standard deviation of mismatch, M, expressed in dB is approximated by
5 . 0
4
21
4
21
2 2 2
22
2
22
2 2
11
2
11
2
2
686 . 8
(
¸
(
¸

.

\

+ I · I +

.

\

+ I +

.

\

+ I =
b a L G b a L b a G
s s s s s s M E(5)
where Γ
G
and Γ
L
are the source and load voltage reflection coefficients respectively and s
11
, s
22
, s
21
are the
scattering coefficients of the twoport component with the suffix a referring to the starting value of the
attenuator and b referring to the finishing value of the attenuator. Harris and Warner
[3]
concluded that the
distribution for M would approximate to that of a normal distribution due to the combination of its
component distributions.
E9.6 The values of Γ
G
and Γ
L
used in Equations E(4) and E(5) and the scattering coefficients used in Equation
E(5) will themselves be subject to uncertainty because they are derived from measurements. This
uncertainty has to be considered when calculating the mismatch uncertainty and it is recommended that
this is done by adding it in quadrature with the measured or derived value of the reflection coefficient; for
example, if the measured value of Γ
L
is 0.03 ± 0.02 then the value of Γ
L
that should be used to calculate
the mismatch uncertainty is
2 2
02 . 0 03 . 0 + , i.e. 0.036.
E10 Directivity
E10.1 When making voltage reflection coefficient (VRC) measurements at rf and microwave frequencies, the
finite directivity of the bridge or reflectometer gives rise to an uncertainty in the measured value of the
VRC, if only the magnitude and not the phase of the directivity component is known. The uncertainty will be
equal to the directivity, expressed in linear terms; e.g. a directivity of 30 dB is equivalent to an uncertainty
of ± 0.0316 VRC.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 41 OF 82
E10.1 As with E9.6 above it is recommended that the uncertainty in the measurement of directivity is taken into
account by adding the measured value in quadrature with the uncertainty, in linear quantities; for example,
if the measured directivity of a bridge is 36 dB (0.016) and has an uncertainty of +8 dB 4 dB (± 0.01) then
the directivity to be used is
2 2
01 . 0 016 . 0 + = 0.019 (34.4 dB).
E11 Test port match
E11.1 The test port match of a bridge or reflectometer used for reflection coefficient measurements will give rise
to an error in the measured VRC due to rereflection. The uncertainty, u(TP), is calculated from u(TP) =
TP.Γ
X
2
where TP is the test port match, expressed as a VRC, and Γ
X
is the measured reflection coefficient.
When a directional coupler is used to monitor incident power in the calibration of a power meter it is the
effective source match of the coupler that defines the value of Γ
G
referred to in E9. As with E9.6 and E10,
the measured value of test port match will have an uncertainty that should be taken into account by using
quadrature summation.
E12 RF connector repeatability
The lack of repeatability of coaxial pair insertion loss and, to a lesser extent, voltage reflection coefficient is
a problem when calibrating devices in a coaxial line measurement system and subsequently using them in
some other system. Although connecting and disconnecting the device can evaluate the repeatability of
particular connector pairs in use, these connector pairs are only samples from a whole population. To
obtain representative data for guidance for various types of connectors in use is beyond the resources of
most measurement laboratories. Reference [5] provides advice on the specifications and use of coaxial
connectors including guidance on the repeatability of the insertion loss of connector pairs.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 42 OF 82
APPENDIX F
SOME SOURCES OF ERROR AND UNCERTAINTY IN MASS CALIBRATIONS
This Appendix describes the more common sources of errors and uncertainties in mass calibration with
brief comments about their nature. They may not all be significant at all levels of measurement, but their
effect should at least be considered when estimating the overall uncertainty of a measurement. Further
information about mass calibration can be found in reference [6].
F1
Reference weight calibration
F1.1 The uncertainties assigned to the values on a calibration certificate for the calibration of the reference
weights are all contributors to the uncertainty budget.
F2 Secular stability of reference weights
F2.1 It is necessary to take into account the likely change in mass of the reference weights since their last
calibration. This change can be estimated from the results of successive calibrations of the reference
weights. If such a history is not available, then it is usual to assume that they may change in mass by an
amount equal to their uncertainty of calibration between calibrations. The stability of weights can be
affected by the material and quality of manufacture (e.g., ill fitting screw knobs), surface finish, unstable
adjustment material, physical wear and damage and atmospheric contamination. The figure adopted for
stability will need to be reconsidered if the usage or environment of the weights changes. The calibration
interval for reference weights will depend on the stability of the weights.
F3 Weighing machine/weighing process
F3.1 The performance of the weighing machine used for the calibration should be assessed to estimate the
contribution it makes to the overall uncertainty of the weighing process. The performance assessment
should cover those attributes of the weighing machine that are significant to the weighing process. For
example, the length of arm error (assuming it is constant) of an equal arm balance need not be assessed if
the weighing process only uses substitution techniques (Borda's method). The assessment may include
some or all of the following:
(a) repeatability of measurement;
(b) linearity within the range used;
(c) digit size/weight value per division, i.e. readability;
(d) eccentricity (off centre load), especially if groups of weights are placed on the weighing pan
simultaneously;
(e) magnetic effects (e.g. magnetic weights, or the effect of force balance motors on cast iron weights);
(f) temperature effects, e.g. differences between the temperature of the weights and the weighing machine;
(f) length of arm error.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 43 OF 82
F4 Air buoyancy effects
F4.1 The accuracy with which air buoyancy corrections can be made depends on how well the density of the
weights is known, and how well the air density can be determined. Some laboratories can determine the
density of weights, but for most mass work assumed figures are used. The air density is usually calculated
from an equation (see reference [6]) after measuring the air temperature, pressure and humidity. For the
highest levels of accuracy, it may also be necessary to measure the carbon dioxide content of the air. The
figures that follow are based upon an air density range of 1.079 kg m
3
to 1.291 kg m
3
which can be
produced by ranges of relative humidity from 30% to 70%, air temperature from 10°C to 30°C and
barometric pressure from 950 millibar to 1050 millibar.
F4.2 For mass comparisons a figure of ± 1 part in 10
6
of the applied mass is typical for common weight
materials such as stainless steel, plated brass, German silver and gunmetal. For cast iron the figure may
be up to ± 3 parts in 10
6
and for aluminium up to ± 30 parts in 10
6
. The uncertainty can be reduced if the
mass comparisons are made within suitably restricted ranges of air temperature, pressure and humidity. If
corrections are made for the buoyancy effects the uncertainty can be virtually eliminated, leaving just the
uncertainty of the correction.
F4.3 Certain weighing machines display mass units directly from the force they experience when weights are
applied. It is common practice to reduce the effects of buoyancy on such devices by the use of an auxiliary
weight, known as a spanning weight, which is used to normalise the readings to the prevailing conditions,
as well as compensating for changes in the machine itself. This spanning weight can be external or internal
to the machine. If such machines are not spanned at the time of use the calibration may be subject to an
increased uncertainty due to the buoyancy effects on the loading weights. For weighing machines that
make use of stainless steel, plated brass, German silver or gunmetal weights this effect may be up to ±16
parts in 10
6
. For cast iron weights the figure may be up to ±18 parts in 10
6
and for aluminium weights up to
± 45 parts in 10
6
.
F4.4 For the ambient conditions stated above the uncertainty limits due to buoyancy effects may be ± 110 parts
in 10
6
and ± 140 parts in 10
6
respectively for comparing water and organic solvents with stainless steel
mass standards, and ± 125 parts in 10
6
and ± 155 parts in 10
6
respectively for direct weighing.
F4.5 Apart from air buoyancy effects, the environment in which the calibration takes place can introduce
uncertainties. Temperature gradients can give rise to convection currents in the balance case, which will
affect the reading, as will draughts from air conditioning units. Rapid changes of temperature in the
laboratory can affect the weighing process. Changes in the level of humidity in the laboratory can make
shortterm changes to the mass of weights, while low levels of humidity can introduce static electricity
effects on some comparators. Dust contamination also introduces errors in calibrations. The movement of
weights during the calibration causes disturbances to the local environment.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 44 OF 82
APPENDIX G
SOME SOURCES OF ERROR AND UNCERTAINTY IN TEMPERATURE CALIBRATIONS
The more common sources of systematic error and uncertainty in the measurement of temperature are
described in this section. Each source may have several uncertainty components.
G1
Reference thermometer calibration
G1.1 The uncertainties assigned to the values on a calibration certificate for the calibration of the reference
thermometer are all contributors to the uncertainty budget.
G2 Measuring instruments
G2.1 The uncertainty assigned to the calibration of any electrical or other instruments used in the
measurements, e.g. standard resistors, measuring bridges and digital multimeters.
G3 Further influences
G3.1 Additional uncertainties in the measurement of the temperature using the reference thermometers:
(a) Drift since the last calibration of the reference thermometers and any associated measuring instruments;
(b) Resolution of reading; this may be very significant in the case of a liquidinglass thermometer or digital
thermometers;
(c) Instability and temperature gradients in the thermal environment, e.g., the calibration bath or furnace,
including any contribution due to difference in immersion of the reference standard from that stated on its
certificate of calibration;
(d) When platinum resistance thermometers are used as reference standards any contribution to the
uncertainty due to selfheating effects should be considered. This will mainly apply if the measuring current
is different from that used in the original calibration and/or the conditions of measurement e.g., `in air' or in
stirred liquid.
G4
Contributions associated with the thermometer to be calibrated
G4.1 These may include factors associated with electrical indicators as well as some of the further influences
already mentioned. When partial immersion liquidinglass thermometers are to be calibrated an additional
uncertainty contribution to account for effects arising from differences in depth of immersion should be
included even when the emergent column temperature is measured.
G4.2 When thermocouples are being calibrated any uncertainty introduced by compensating leads and
reference junctions should be taken into account. Similarly any thermal emfs introduced by switches or
scanner units should be investigated. Unknown errors arising from inhomogeneity of the thermocouple
being calibrated can give rise to significant uncertainties. Ideally this should be evaluated at the time of
calibration, possibly by varying the immersion depth of the thermocouple in an isothermal enclosure. For
many calibrations, however, this will not be practical. In such cases, a figure of 20% of the maximum
permissible error for the particular thermocouple type is considered reasonable.
G5 Mathematical interpretation
G5.1 Uncertainty arising from mathematical interpretation, e.g. in applying scale corrections or deviations from a
reference table, or in curvefitting to allow for scale nonlinearity, should be assessed.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 45 OF 82
APPENDIX H
SOME SOURCES OF ERROR AND UNCERTAINTY IN DIMENSIONAL CALIBRATIONS
The more common sources of systematic error and uncertainty in dimensional measurements are
described in this section.
H1 Reference standards and Instrumentation
H1.1 The uncertainties assigned to the reference standards and those for the measuring instruments used to
make the measurements.
H2 Secular stability of reference standards and instrumentation
H2.1 The changes that occur over time must be taken into account, usually by reference to the calibration
history of the equipment. This is particularly important when the equipment may be exposed to physical
wear as part of normal operation.
H3 Temperature effects
H3.1 The uncertainties associated with differences in temperature between the gauge being calibrated and the
reference standards and measuring instruments used should be accounted for. These will be most
significant over the longer lengths and in cases involving dissimilar materials. Whilst it may be possible to
make corrections for temperature effects there will be residual uncertainties resulting from uncertainty in
the values used for the coefficients of expansion and the calibration of the measuring thermometer.
H4 Elastic compression
H4.1 These are uncertainties associated with differences in elastic compression between the materials from
which the gauge being calibrated and the reference standards were manufactured. They are likely to be
most significant in the more precise calibrations and in cases involving dissimilar materials. They will relate
to the measuring force used and the nature of stylus contact with the gauge and reference standard. Whilst
mathematical corrections can be made there will be residual uncertainties resulting from the uncertainty of
the measuring force and in the properties of the materials involved.
H5 Cosine errors
H5.1 Any misalignment of the gauge being calibrated or reference standards used, with respect to the axis of
measurement, will introduce errors into the measurements. Such errors are often referred to as cosine
errors and can be minimised by adjusting the attitude of the gauge with respect to the axis of measurement
to find the relevant turning points that give the appropriate maximum or minimum result. Small residual
errors can still result where, for instance, incorrect assumptions are made concerning any features used for
alignment of the datums.
H6 Geometric errors
H6.1 Errors in the geometry of the gauge being calibrated, any reference standards used or critical features of
the measuring instruments used to make the measurements can introduce additional uncertainties.
Typically these will include small errors in the flatness or sphericity of stylus tips, the straightness, flatness,
parallelism or squareness of surfaces used as datum features, and the roundness or taper in cylindrical
gauges and reference standards. Such errors are often most significant in cases where perfect geometry
has been wrongly assumed and where the measurement methods chosen do not capture, suppress or
otherwise accommodate the geometric errors that prevail in a particular case.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 46 OF 82
APPENDIX J
SOME SOURCES OF ERROR AND UNCERTAINTY IN PRESSURE CALIBRATIONS
USING DEAD WEIGHT TESTERS
The more common sources of systematic error and uncertainty in the generation of known pressures,
using dead weight testers (DWT), are described in this section.
J1 Reference DWT
J1.1 The uncertainties assigned to the values on a calibration certificate for the reference dead weight tester
are all contributors to the uncertainty budget. These include the following:
(a) Area uncertainty including any uncertainty in the distortion. This uncertainty will often vary with pressure;
(b) Piston and weight carrier mass.
J2 Secular stability of the reference dead weight tester
J2.1 It is necessary to account for likely changes in the area and mass of the reference DWT since the last
calibration. This change can be estimated from successive calibrations of the reference DWT. The
secular stability uncertainty for the area will depend on the calibration interval and can be larger than the
calibration uncertainty. The variation between calibrations in the area of a DWT will depend on its usage,
design, and material composition and is therefore a best estimate from actual data. Where this is not
available it is recommended that a pessimistic estimate is made and a short calibration interval set.
J2.2 The drift of the piston mass will be larger in oil DWTs as this will reflect the difficulties in repeat weighting
of pistons that have been immersed in oil. These difficulties arise from incomplete cleaning processes
and possible instability due to the evaporation of solvents.
J3 Reference DWT mass set uncertainty
J3.1 The uncertainties assigned to the values on a calibration certificate for the weights in the reference dead
weight tester mass set are all contributors to the uncertainty budget. The uncertainty of the mass stack
used to generate pressure should be evaluated over the range of the DWT. The relative uncertainty is
often higher at lower pressures.
J4 Secular stability of the reference DWT mass set
J4.1 It is necessary to account for likely changes the mass set of the reference DWT since the last
calibration. Paragraph F2.1 addresses the subject of secular stability of reference weights.
J5 Uncertainty of local gravity determination
J5.1 The pressure generated by a DWT is directly affected by the local acceleration due to gravity, g. With
care, this can be measured with an uncertainty of less than 1 ppm. It is possible for an estimate of the g
value to be obtained from a reputable geological survey organisation based on a grid reference; this
would attract an uncertainty of around 3 ppm. It can also be calculated from knowledge of latitude and
altitude, however the uncertainty will be much larger  around 50 ppm in the UK. Some knowledge of the
Bouguer anomalies is required to achieve these levels of uncertainty from such calculations.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 47 OF 82
J6 Air buoyancy effect
J6.1 Air buoyancy affects the mass set of a DWT in the same way as described in paragraph F4.
J7 Temperature effect on DWT area
J7.1 The area on a DWT changes with temperature; its temperature coefficient of expansion being related to
the particular materials that the piston and cylinder are made from. Consideration has to be given to any
variation in temperature from the reference temperature when the DWT was calibrated, variation in
temperature during a calibration and uncertainty in the determination of the piston temperature.
J8 Uncertainty due to head correction
J8.1 Any difference between the height of the reference DWT datum level and that of the item being
calibrated will affect the pressure generated at that item. For pneumatic calibrations this effect is
proportional to pressure and normally equates to about 116 ppm/m. For hydraulic calibrations the effect
is a fixed pressure effect that will depend on the density of the fluid used, local acceleration due to
gravity and the height difference. (Fluid head pressure = ρ.g.h) For most DWT oils the effect is between
8 and 9 Pa/mm.
J8.2 The float height position of the piston will also contribute to the head correction uncertainty. This effect
will be related to the fall rate of the piston and the particular measurement procedure in use.
J9 Effects of fluid properties
J9.1 For hydraulic calibrations the effect of the fluid properties on fluid head corrections, buoyancy volume
corrections and surface tension corrections will also need to be considered. These figures are usually
reported on calibration certificates for DWTs. However, care must be taken to convert any quoted
correction to the actual oil used if different from that used during the calibration of the reference DWT. In
most circumstances the uncertainty of these influence quantities can be treated as negligible.
J10 Nonverticality of the DWT piston
An uncertainty arises due to the fact that the piston may not be perfectly vertical. If it were, then all of the
force would act on the area. Any departure from vertical will reduce the force and therefore the generated
pressure. The effect in terms of generated pressure is proportional to the cosine of the angle from true
vertical.
J11 Uncertainties arising from the calibration process
J11.1 Any uncertainty arising from the calibration process will need to be evaluated. These could include the
resolution and repeatability of the unit being calibrated and the effects of the environment on it.
Uncertainties due to calculation or data fitting of the calibration results may also have to be considered.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 48 OF 82
APPENDIX K
EXAMPLES OF APPLICATION FOR CALIBRATION
NOTES
(a) This Appendix presents a number of example uncertainty budgets in various fields of measurement. The
examples are not intended as preferred or mandatory requirements. They are presented to illustrate the
principles involved in uncertainty evaluation and to show how the common sources of uncertainty in the
various fields can be analysed in practice. They are, however, believed to be realistic for the particular
measurements described.
(b) These examples may also be used for the purpose of software validation. If an uncertainty budget has
been prepared using a spreadsheet, the configuration of the spreadsheet can be verified by entering the
same values and comparing the output of the spreadsheet with the results shown in the examples.
K1 Calibration of a 10 kΩ resistor by voltage intercomparison
K1.1 Ahighresolution digital voltmeter is used to measure the voltages developed across a standard resistor
and an unknown resistor of the same nominal value as the standard, when the seriesconnected resistors
are supplied from a constant current dc source. Both resistors are immersed in a temperature controlled oil
bath maintained at 20.0°C. The value of the unknown resistor, R
X
, is given by
( ) ( ) ( )
S
X
TC D S X
V
V
V s t R R R R · + · + + = Δ δ , where
R
S
δR
D
R
TC
Δt
V
X
V
S
( ) V s
=
=
=
=
=
=
=
Calibration value for the standard resistor,
Relative drift in R
S
since the previous calibration,
Relative temperature coefficient of resistance for R
S
,
Maximum variation in oil bath from nominal temperature,
Voltage across R
X
,
Voltage across R
S
,
Repeatability of ratio V
X
/V
S
.
K1.2 The calibration certificate for the standard resistor reported an uncertainty of ± 0.5 ppm at a coverage
probability of approximately 95% (k = 2).
K1.3 Acorrection was made for the estimated drift in the value of R
S
. The uncertainty in this correction, R
D
, was
estimated to have limits of ± 0.5 ppm.
K1.4 The temperature coefficient of resistance for the standard resistor was obtained from a graph of
temperature versus resistance. Such curves are normally parabolic in nature, however using a linear
approximation over the small range of temperature variation encountered in the bath, a value of ± 2.5 ppm
per ºC was assigned. This value was included in the uncertainty budget as a sensitivity coefficient.
K1.5 Records of evaluation of the oil bath characteristics showed that the maximum temperature deviation from
the set point did not exceed ± 0.1ºC at any point within the bath.
K1.6 The same voltmeter is used to measure V
X
and V
S
and although the uncertainty contributions will be
correlated the effect is to reduce the uncertainty and it is only necessary to consider the relative difference
in the voltmeter readings due to linearity and resolution, which was estimated to have limits of ±0.2 ppm for
each reading. Each of these is assigned a rectangular distribution.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 49 OF 82
K1.7 Type Aevaluation: Five measurements were made to record the departure from unity in the ratio V
X
/V
S
in
ppm. The readings were as follows:
+10.4, +10.7, +10.6, +10.3, +10.5
From Equation (3), the mean value V = +10.5 ppm.
From Equations (5) and (6), ( ) ( )
5
158 . 0
= = V s V u = 0.0707 ppm.
K1.8 Uncertainty budget
Symbol Source of uncertainty
Value
±
Probability
distribution
Divisor
c
i
u
i
(R
X
)
ppm
ν
i
or
ν
eff
RS
Calibration of standard resistor
0.5 ppm Normal 2 1 0.25 ∞
δR
D
Uncorrected drift since last calibration 0.5 ppm Rectangular √3 1 0.289
∞
Δt Temperature effects 0.1 ºC Rectangular √3 2.5 ppm/ºC 0.144 ∞
V
S
Voltmeter across R
S 0.2 ppm Rectangular √3 1 0.115 ∞
V
X
Voltmeter across R
X 0.2 ppm Rectangular √3 1 0.115 ∞
u(V)
Repeatability of indication
0.071 ppm Normal 1 1 0.071 4
u
c
(y) Combined standard uncertainty Normal 0.445 >500
U Expanded uncertainty
Normal
(k = 2)
0.891 >500
K1.9 Reported result
The measured value of the 10 kΩ resistor at 20 ºC ± 0.1º C was 10 000.105 Ω ± 0.89 ppm
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2, providing
a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in accordance with
UKAS requirements.
NOTE
The temperature coefficient of the resistor being calibrated is normally not included, as it is an unknown quantity. The relevant
temperature conditions are therefore included in the reporting of the result.
K2 Calibration of a coaxial power sensor at a frequency of 18 GHz
K2.1 The measurement involves the calibration of an unknown power sensor against a standard power
sensor by substitution on a stable, monitored source of defined source impedance. The measurement is
made in terms of Calibration Factor, defined as
frequency n calibratio at power Incident
frequency reference at power Incident
, for the same
power sensor response and is determined from the following:
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 50 OF 82
Calibration Factor, K
X
= (K
S
+ D
S
) x δDC x δM x δREF, where
K
S
= Calibration Factor of the standard sensor,
D
S
= Drift in standard sensor since the previous calibration,
δDC
= Ratio of DC voltage outputs,
δM = Ratio of Mismatch Losses,
δREF = Ratio of reference power source (shortterm stability of 50 MHz reference).
K2.2 Four separate measurements were made which involved disconnection and reconnection of both the
unknown sensor and the standard sensor on a power transfer system. All measurements were made in
terms of voltage ratios that are proportional to calibration factor.
K2.3 None of the uncertainty contributions are considered to be correlated to any significant extent.
K2.4 There will be mismatch uncertainties associated with the source/standard sensor combination and with
the source/unknown sensor combination. These will be 200Γ
G
Γ
S
% and 200Γ
G
Γ
X
% respectively, where
Γ
G
= 0.02 at 50 MHz and 0.07 at 18 GHz,
Γ
S
= 0.02 at 50 MHz and 0.10 at 18 GHz,
Γ
X
= 0.02 at 50 MHz and 0.12 at 18 GHz.
These values include the uncertainty in the measurement of Γ as described in paragraph E9.6.
K2.5 The standard power sensor was calibrated by an accredited laboratory 6 months before use; the expanded
uncertainty of ± 1.1% was quoted for a coverage factor k = 2.
K2.6 The longterm stability of the standard sensor was estimated from the results of 5 annual calibrations. No
predictable trend could be detected so drift corrections could not be made. The error due to secular
stability was therefore assumed to be zero with limits, in this case, not greater than ± 0.4% per year. A
value of ± 0.2% was used as the previous calibration was within 6 months.
K2.7 The instrumentation linearity uncertainty was estimated from measurements against a reference
attenuation standard. The expanded uncertainty for k = 2 of ± 0.1% applies to ratios up to 2:1.
K2.8 Type A evaluation: The four measurements resulted in the following values of Calibration Factor:
93.45% 92.20% 93.95% 93.02%
From Equation (3), the mean value
X
K = 93.16%.
From Equations (5) and (6), ( ) ( )
4
7415 . 0
= =
X R
K s K u = 0.3707%.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 51 OF 82
K2.9 Uncertai nty budget
Symbol Source of uncertainty
Value
± %
Probability
distribution
Divisor c
i
u
i
(K
X
)
%
ν
i
or
νeff
K
S
Calibration factor of standard
1.1 Normal 2.0 1.0 0.55 ∞
D
S
Drift since last calibration 0.2 Rectangular √3 1.0 0.116
∞
δDC
Instrumentationlinearity 0.1 Normal 2.0 1.0 0.05 ∞
δM
Stability of 50 MHz reference
0.2 Rectangular √3 1.0 0.116 ∞
M
1
M
2
M
3
M
4
Mismatch:
Standard sensor at 50 MHz
Unknown sensor at 50 MHz
Standard sensor at 18 GHz
Unknown sensor at 18 GHz
0.08
0.08
1.40
1.68
Ushaped
Ushaped
Ushaped
Ushaped
√2
√2
√2
√2
1.0
1.0
1.0
1.0
0.06
0.06
0.99
1.19
∞
∞
∞
∞
K
R
Repeatability of indication 0.37 Normal 1.0 1.0 0.37 3
u(K
X
) Combined standard uncertainty Normal 1.69 >500
U Expanded uncertainty
Normal
(k = 2)
3.39 >500
K2.10 Reported result
The measured calibration factor at 18 GHz was 93.2 % ± 3.4 %.
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2, providing
a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in accordance with
UKAS requirements.
NOTES
1 For the measurement of calibration factor, the uncertainty in the absolute value of the 50 MHz reference source need not be
included if the standard and unknown sensors are calibrated using the same source, within the timescale allowed for its short
term stability.
2 This example illustrates the significance of mismatch uncertainty in measurements at relatively high frequencies.
3 In a subsequent use of a sensor further random components of uncertainty may arise due to the use of different connector pairs.
K3 Calibration of a 30 dB coaxial attenuator
K3.1 The measurement involves the calibration of a coaxial step attenuator at a frequency of 10 GHz using a
dual channel 30 MHz IF substitution measurement system. The measurement is made in terms of the
attenuation in dB between a matched source and load from the following:
A
X
= A
b
– A
a
+ A
IF
+ D
IF
+ L
M
+ R
D
+ M + A
L
+ A
R
, where
A
b
= Indicated attenuation with unknown attenuator set to zero,
A
a
= Indicated attenuation with unknown attenuator set to 30 dB,
A
IF
= Calibration of reference IF attenuator,
D
IF
= Drift in reference IF attenuator since last calibration,
L
M
= Departurefrom linearity of mixer,
R
D
= Error due to resolution of detection system,
M = Mismatch error,
A
L
= Effect of signal leakage,
A
R
= Repeatability.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 52 OF 82
K3.2 The result is corrected for the calibrated value of the IF attenuator using the results from a calibration
certificate, which gave an uncertainty of ± 0.005 dB at a coverage probability of 95% (k = 2).
K3.3 No correction is made for the drift of the IF attenuator. The limits of ± 0.002 dB were estimated from the
results of previous calibrations.
K3.4 No correction is made for mixer nonlinearity. The uncertainty was estimated from a series of linearity
measurements over the dynamic range of the system to be ± 0.002 dB/10dB. An uncertainty of ± 0.006 dB
was therefore assigned at 30 dB. The probability distribution is assumed to be rectangular.
K3.5 The resolution of the detection system was estimated to cause possible rounding errors of onehalf of one
least significant recorded digit i.e. ± 0.005 dB. This occurs twice  once for the 0 dB reference setting and
again for the 30 dB measurement. Two identical rectangular distributions with semirange limits of a
combine to give a triangular distribution with semirange limits of 2a. The uncertainty due to resolution is
therefore 0.01 dB with a triangular distribution.
K3.6 No correction is made for mismatch error. The mismatch uncertainty is calculated from the scattering
coefficients using Equation E(5). The values used were as follows:
Γ
L
= 0.03 Γ
G
= 0.03 s
11a
= 0.05 s
11b
= 0.09 s
22a
= 0.05 s
22b
= 0.01 s
21a
= 1 s
21b
= 0.031
K3.7 Special experiments were performed to determine whether signal leakage had any significant effect on
the measurement system. No effect greater than ± 0.001 dB could be observed for attenuation values up
to 70 dB. The probability distribution is assumed to be rectangular.
K3.8 Type A evaluation: Four measurements were made which involved setting the reference level with the
step attenuator set to zero and then measuring the attenuation for the 30 dB setting. The results were as
follows:
30.04 dB 30.07 dB 30.03 dB 30.06 dB
From Equation (3), the mean value
X
A = 30.050 dB.
From Equations (5) and (6), ( ) ( )
4
018 . 0
= =
X R
A s A u = 0.009 dB.
K3.9 Uncertainty budget
Symbol Source of uncertainty
Value
± dB
Probability
distribution
Divisor c
i
u
i
(A
X
)
dB
ν
i
or
ν
eff
A
IF
Calibration of reference attenuator
0.005 Normal 2.0 1.0 0.0025 ∞
D
IF
Drift since last calibration 0.002 Rectangular √3 1.0 0.0016
∞
L
M
Mixer nonlinearity 0.006 Rectangular √3 1.0 0.0035 ∞
R
D
Resolution of indication
0.010 Triangular √6 1.0 0.0041 ∞
M
Mismatch
0.022 Normal 1 1.0 0.022 ∞
A
L
Signal leakage effects 0.001 Rectangular √3 1.0 0.0006 ∞
A
R
Repeatability of indication 0.009 Normal 1.0 1.0 0.009 3
u(A
X
) Combined standard uncertainty Normal 0.0246 >500
U Expanded uncertainty
Normal
(k = 2)
0.0491 >500
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 53 OF 82
K3.10 Reported result
The measured value of the 30 dB attenuator at 10 GHz was 30.050 dB ± 0.049 dB.
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2,
providing a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in
accordance with UKAS requirements.
NOTES
1 Combination of relatively small uncertainties expressed in dB is permissible since log
e
(1+x) ≈ x when x is small and
2.303log
10
(1+x) ≈ x. For example: 0.1 dB corresponds to a power ratio of 1.023 and 2.303log
10
(1+0.023) = 0.0227.
Thus relatively small uncertainties expressed in dB may be combined in the same way as those expressed as linear relative
values, e.g. percentage.
2 For attenuation measurements, the probability distribution for RF mismatch uncertainty is dependent on the combination of at
least three mismatch uncertainties and can be treated as having a normal distribution. For further details see paragraph
E9.5.
3 In a subsequent use of an attenuator further random components of uncertainty may arise due to the use of different connector
pairs.
K4 Calibration of a weight of nominal value 10 kg of OIML Class M1
K4.1 The calibration is carried out using a mass comparator whose performance characteristics have
previously been determined, and a weight of OIML Class F2. The unknown weight is obtained from:
Measured value of unknown weight, W
X
= W
S
+ D
S
+ δI
d
+ δC + W
R
+ A
b
, where
W
S
= Weight of the standard,
D
S
= Drift of standard since last calibration,
δI
d
= The rounding of the value of the least significant digit of the indication,
δC = Difference in comparator readings,
W
R
= Repeatability,
A
b
= Correction for air buoyancy.
K4.2 The calibration certificate for the standard mass gives an uncertainty of ± 30 mg at a coverage
probability of approximately 95% (k = 2).
K4.3 The monitored drift limits for the standard mass have been set equal to the expanded uncertainty of its
calibration, and are ± 30 mg. A rectangular probability distribution has been assumed.
K4.4 The least significant digit I
d
for the mass comparator represents 10 mg. Digital rounding δI
d
has limits of
± 0.5I
d
for the indication of the values of both the standard and the unknown weights. Combining these
two rectangular distributions gives a triangular distribution, with uncertainty limits of ± I
d
, that is ± 10 mg.
K4.5 The linearity error of the comparator over the 2.5 g range permitted by the laboratory's procedures for
the comparison was estimated from previous measurements to have limits of ± 3 mg. A rectangular
probability distribution has been assumed.
K4.6 A previous Type A evaluation of the repeatability of the measurement process, comprising 10
comparisons between standard and unknown, gave a standard deviation, s(W
R
), of 8.7 mg. This
evaluation replicates the normal variation in positioning single weights on the comparator, and therefore
includes effects due to eccentricity errors.
K4.7 No correction is made for air buoyancy, for which limits were estimated to be ± 1 ppm of nominal value
i.e. ± 10 mg. A rectangular probability distribution has been assumed.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 54 OF 82
K4.8 Three results were obtained for the unknown weight using the conventional technique of bracketing the
reading with two readings for the standard. The results were as follows:
No. Weight on pan Comparator
reading
Standard mean unknown  standard
standard + 0.01 g
1 unknown + 0.03 g + 0.015 g + 0.015 g
standard + 0.02 g
2 unknown + 0.04 g + 0.015 g + 0.025 g
standard + 0.01 g
3 unknown + 0.03 g + 0.010 g + 0.020 g
standard + 0.01 g
Mean difference: + 0.020 g
From the calibration certificate, the mass of the standard is 10 000.005 g. The calibrated value of the
unknown is therefore 10 000.005 g + 0.020 g = 10 000.025 g.
K4.9 Since three comparisons between standard and unknown were made (using 3 readings on the unknown
weight), this is the value of n that is used to calculate the standard deviation of the mean:
From Equations (5) and (6), ( ) ( )
3
7 . 8
= =
X R
W s W u = 5.0 mg.
K4.10 Uncertainty budget
Symbol Source of uncertainty
Value
± mg
Probability
distribution
Divisor c
i
u
i
(W
X
)
mg
ν
i
or
ν
eff
W
S
Calibration of standard weight 30.0 Normal 2.0 1.0 15.0 ∞
D
S
Drift since last calibration 30.0 Rectangular √3 1.0 17.32 ∞
δI
d
Digital rounding error, comparison
10.0 Triangular √6 1.0 4.08 ∞
δC
Comparator nonlinearity
3.0 Rectangular √3 1.0 1.73 ∞
A
b
Air buoyancy (1 ppm of nominal value) 10.0 Rectangular √3 1.0 5.77 ∞
W
R
Repeatability of indication 5.0 Normal 1.0 1.0 5.0 9
u(W
X
) Combined standard uncertainty Normal 24.55 >500
U Expanded uncertainty
Normal
(k = 2)
49.11 >500
K4.11 Reported result
The measured value of the 10 kg weight was 10 000.025 g ± 0.049 g.
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2,
providing a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in
accordance with UKAS requirements.
NOTE
The degrees of freedom shown in the uncertainty budget are derived from a previous evaluation of repeatability, for which 10
readings were used (see paragraph B4).
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 55 OF 82
K5 Calibration of a weighing machine of 205 g capacity by 0.1 mg digit
K5.1 The calibration is carried out using weights of OIML Class E2. Checks will normally be carried out for
linearity of response across the nominal capacity of the weighing machine, eccentricity effects of the
positioning of weights on the load receptor, and repeatability of the machine for repeated weighings near
full load. The span of the weighing machine has been adjusted using its internal weight before calibration.
The following uncertainty calculation is carried out for a near full loading of 200 g. The machine indications
are obtained from
Unknown indication I
X
= W
S
+ D
S
+δI
d0
+ δI
d
+ A
b
+ I
R
, where
W
S
= Weight of the standard,
D
S
= Drift of standard since last calibration,
δI
d0
= The rounding of the value of one digit at the zero reading,
δI
d
= The rounding of the value of one digit of the indicated value,
A
b
= Correction for air buoyancy,
I
R
= Repeatability of the indication.
K5.2 The calibration certificate for the stainless steel 200 g standard mass gives an uncertainty of ± 0.1 mg at a
coverage probability of approximately 95% (k = 2).
K5.3 No correction is made for drift, but the calibration interval is set so as to limit the drift to ± 0.1 mg. The
probability distribution is assumed to be rectangular.
K5.4 It is often the case that when a weighing machine is zeroed, or tared, it does so to a greater resolution than
that provided by the digital readout. This can be thought of as an “internal digit” that is not presented to the
user. In this example, it is assumed that the “internal digit” size is onetenth of that provided on the external
display. The resulting possible rounding error is therefore ± 0.005 mg. The probability distribution is
assumed to be rectangular.
NOTE
This is a good example of how a detailed knowledge of the principles of operation of a measuring instrument may be
required in order to identify and quantify the associated uncertainties.
K5.5 No correction can be made for the rounding due to the resolution of the digital display of the machine. The
least significant digit on the range being calibrated corresponds to 0.1 mg and there is therefore a possible
rounding error of ± 0.05 mg. The probability distribution is assumed to be rectangular.
K5.6 No correction is made for air buoyancy. As the span of the machine was adjusted with its internal weight
before calibration, the uncertainty limits were estimated to be ± 1 ppm of the nominal value, i.e. ± 0.2 mg.
K5.7 The repeatability of the machine was established from a series of 10 readings (Type Aevaluation), which
gave a standard deviation, s(W
R
), of 0.05 mg. The degrees of freedom for this evaluation are 9, i.e. n  1.
K5.8 Only one reading was taken to establish the weighing machine indication for each linearity and eccentricity
point. For this calibration point the weighing machine indication, I
X
, was 199.9999 g when the 200 g
standard mass was applied. The value of n that is used to calculate the standard deviation of the mean of
the indication, using the previously obtained repeatability standard deviation, s(W
R
), is therefore one:
From Equations (5) and (6), ( ) ( )
1
05 . 0
= =
X R
W s I u = 0.05 mg.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 56 OF 82
K5.9 Uncertainty budget
Symbol Source of uncertainty
Value
± mg
Probability
distribution
Divisor c
i
u
i
(I
X
)
mg
ν
i
or
νeff
W
S
Calibration of standard weight
0.1 Normal 2.0 1.0 0.05 ∞
D
S
Drift since last calibration 0.1 Rectangular √3 1.0 0.058
∞
δI
d0
Digital rounding error (at zero) 0.005 Rectangular √3 1.0 0.003 ∞
δI
d
Digital rounding error (for indicated value)
0.05 Rectangular √3 1.0 0.029 ∞
A
b
Air buoyancy (1 ppm of nominal value) 0.2 Rectangular √3 1.0 0.115 ∞
I
R
Repeatability of indication
0.05 Normal 1.0 1.0 0.05 9
u(I
X
) Combined standard uncertainty Normal 0.150 >500
U Expanded uncertainty
Normal
(k = 2)
0.300 >500
K5.10 Reported result
For an applied weight of 200 g the indication of the weighing machine was 199.9999 g ± 0.30 mg.
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2, providing
a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in accordance with
UKAS requirements.
K6 Calibration of a Grade 2 gauge block of nominal length 10 mm
K6.1 The calibration was carried out using a comparator with reference to a grade K standard gauge block of
similar material. The length of the unknown gauge block, L
X
, was determined from
L
X
= L
S
+ L
D
+ δL – [L(αδt + δαδT)] + D
C
+ δC + L
V(x)
+ L
r
, where
L
S
= Certified length of the standard gauge block at 20ºC
L
D
= Drift with time of certified length of standard gauge block
δL = Measured difference in length
L = Nominal length of gauge block
α = Mean thermal expansion coefficient of the standard and unknown gauge blocks
δt = Difference in temperature between the standard and unknown gauge blocks
δα = Difference in thermal expansion coefficients of the standard and unknown gauge blocks
δT = Difference in mean temperature of gauge blocks and reference temperature of 20ºC when δL is
determined
D
C
= Discrimination and linearity of the comparator
δC = Difference in coefficient of compression of standard and unknown gauge blocks
L
V(x)
= Variation in length with respect to the measuring faces of the unknown gauge block
L
r
= Repeatability of measurement
K6.2 The value of L
S
was obtained from the calibration certificate for the standard gauge block. The associated
uncertainty was ± 0.03 μm (k = 2).
K6.3 The change in value L
D
of the standard gauge block with time was estimated from previous calibrations to
be zero with an uncertainty of ± 15 nm. From experimental evidence and prior experience the value of zero
was considered the most likely, with diminishing probability that the value approached the limits. A
triangular distribution was therefore assigned to this uncertainty contribution.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 57 OF 82
K6.4 The coefficient of thermal expansion applicable to each gauge block was assumed to have a value, α, of
11.5 μm m
1
ºC
1
with limits of ± 1 μm m
1
ºC
1
. Combining these two rectangular distributions, the difference
in thermal expansion coefficient between the two blocks, δα, is ± 2 μm m
1
ºC
1
with a triangular distribution.
For L = 10 mm this corresponds to ± 20 nm/ºC. This difference will have two influences:
(a)
The difference in temperature, δt, between the standard and unknown gauge blocks was estimated to be
zero with limits of ± 0.08 ºC, giving rise to a length uncertainty of ± 1.6 nm.
(b) The difference δT between the mean temperature of the two gauge blocks and the reference temperature
of 20 ºC was measured and was assigned limits of ± 0.2 ºC, giving rise to a length uncertainty of ± 4 nm.
As the influence of δα appears directly in both these uncertainty contributions they are considered to be
correlated and, in accordance with paragraph D3.3, the corresponding uncertainties have been added
before being combined with the remaining contributions. This is included in the uncertainty budget as δ
Ts,x
.
K6.5
The error due to the discrimination and nonlinearity of the comparator, D
C
, was taken as zero with limits of
± 0.05 μm, assessed from previous measurements. Similarly, the difference in elastic compression δC
between the standard and unknown gauge blocks was estimated to be zero with limits of ± 0.005 μm.
K6.6 The variation in length of the unknown gauge block, L
V(x)
, was considered to comprise two components:
(i) Effect due to incorrect central alignment of the probe; assuming this misalignment was within a circle of
radius 0.5 mm, calculations based on the specifications for grade C gauge blocks indicted an uncertainty of
± 17 nm. (ii) Effects due to surface irregularities such as scratches or indentations; such effects have a
detection limit of approximately 25 nm when examined by experienced staff. Quadrature combination of
these contributions gives an uncertainty due to surface irregularities of ± 30 nm. As this is the combination
of two rectangular distributions, of similar magnitude, the resulting distribution was assumed to be
triangular.
K6.7 The repeatability of the calibration process was established from previous measurements using gauge
blocks of similar type and nominal length. This Type Aevaluation, based upon 11 measurements and using
Equation (7), yielded a standard deviation s(L
R
) of 16 nm.
K6.8 The calibration of the unknown gauge block was established from a single measurement; however, as the
conditions were the same as for the previous evaluation of repeatability the standard uncertainty due to
repeatability can be obtained from this previous value of standard deviation, with n = 1, because only one
reading is made for the actual calibration.
From Equation (6), ( ) ( )
1
16
= =
X R
L s L u = 16 nm.
The measured length of the unknown gauge block was 9.999 94 mm.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 58 OF 82
K6.9 Uncertainty budget
Symbol Source of uncertainty
Value
± nm
Probability
distribution
Divisor c
i
u
i
(L
X
)
nm
ν
i
or
ν
eff
L
S
Calibration of the standard gauge block
30 Normal 2.0 1.0 15.0
∞
L
D
Drift since last calibration 15 Triangular √6 1.0 6.1 ∞
D
C
Comparator 50 Rectangular √3 1.0 28.9 ∞
δC Difference in elastic compression 5.0 Rectangular √3 1.0 2.9 ∞
δT
s,x
Temperature effects 5.6 Triangular √6 1.0 2.3 ∞
L
V(x)
Length variation of unknown gauge block 30 Triangular √6 1.0 12.3 ∞
L
r
Repeatability 16 Normal 1.0 1.0 16.0 10
u(L
X
)
Combined standard uncertainty
Normal 39.0 >400
U
Expanded uncertainty
Normal
(k = 2)
77.9 >400
K6.10 Reported result
The measured length of the gauge block was 9.999 94 mm ± 0.078 μm.
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2,
providing a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in
accordance with UKAS requirements.
K7 Calibration of a Type N thermocouple at 1000ºC
K7.1 AType N thermocouple is calibrated against two reference standard Type R thermocouples in a
horizontal furnace at a temperature of 1000ºC. The emfs generated by the thermocouples are measured
using a digital microvoltmeter via a selector/reversing switch. All the thermocouples have their reference
junctions at 0ºC. The unknown thermocouple is connected to the reference point using compensating
cables.
K7.2 The temperature t
x
of the hot junction of the unknown thermocouple is given by
F D
S
S
R iS iS iS s x
t t
C
t
V V V V t t δ δ
δ
δ δ δ + +


.

\

÷ + + + =
0
0
2 1
F D S
S
s
R s iS s iS s iS s
t t t
C
C
V C V C V C V t δ δ δ δ δ δ + + ÷ · + · + · + · ~
0
0
2 1
The voltage V
x
(t) across the thermocouple wires with the reference junction at 0 ºC during the calibration
is
( ) ( )
0
0
2 1
0
0
X
X
X
R TH iX iX iX
X
X
x
x x x
C
t
C
t
V h V V V
C
t
C
t
t V t V
o
o o o
o
÷
A
+ + + + + = ÷
A
+ ~
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 59 OF 82
where
t
S
(V) = Temperature of the reference thermometer in terms of voltage with the cold junction
at 0 ºC. The function is given in the calibration certificate.
V
iS
, V
iX
= Indication of the microvoltmeter.
δV
iS1
, δV
iX1
= Voltage corrections due to the calibration of the microvoltmeter.
δV
iS2
, δV
iX2
= Rounding errors due to the resolution of the microvoltmeter.
δV
R
= Voltage error due to contact effects of the reversing switch.
δt
0S
, δt
0X
= Temperature corrections associated with the reference junctions.
C
S
, C
X
= Sensitivity coefficients of the thermocouples for voltage at the measurement
temperature of 1000 ºC.
C
S0
, C
X0
= Sensitivity coefficients of the thermocouples for voltage at the reference temperature
of 0 ºC.
δt
D
= Drift of the reference thermometers since the last calibration.
δt
F
= Temperature correction due to nonuniformity of the furnace.
t
= Temperature at which the unknown thermocouple is to be calibrated.
Δt = t  t
X
= Deviation of the temperature of the calibration point from the temperature of the
furnace.
δV
LX
= Voltage error due to the compensation leads.
h
TH
= Error due to inhomogeneity of the unknown thermocouple.
K7.3 The reported result is the output emf of the test thermocouple at the temperature of the hot junction. The
measurement process consists of two parts  determination of the temperature of the furnace and
determination of the emf of the test thermocouple The evaluation of uncertainty has therefore been split
into two parts to reflect this situation.
K7.4 The Type R reference thermocouples are supplied with calibration certificates that relate the temperature
of their hot junctions with their cold junctions at 0 ºC to the voltage across their wires. The expanded
uncertainty U is ± 0.3 ºC with a coverage factor k = 2.
K7.5 No correction is made for drift of the reference thermocouples since the last calibration but an
uncertainty of ± 0.3ºC has been estimated from previous calibrations. A rectangular probability
distribution has been assumed.
K7.6 The voltage sensitivity coefficients of the reference and unknown thermocouples have been obtained
from reference tables as follows:
Sensitivity coefficient at temperatures of
Thermocouple
0 ºC 1000 ºC
reference
C
S
= 0.189 ºC/μV C
S0
= 0.077 ºC/μV
unknown C
X
= 0.038 ºC/μV C
X0
= 0.026 ºC/μV
K7.7 The least significant digit of the microvoltmeter corresponds to a value of 1 μV. This results in possible
rounding errors, δV
iS2
and δV
iX2
, of ± 0.5 μV for each indication.
K7.8 Corrections were made to the microvoltmeter readings by using data from its calibrati on certificate. Drift
and other influences were all considered negligible, therefore only the calibration uncertainty of ± 2.0 μV
(k = 2) is to be included in the uncertainty budget.
K7.9 Residual parasitic offset voltages due to the switch contacts were estimated to be zero within ± 2.0 μV.
K7.10 The temperature of the reference junction of each thermocouple is known to be 0 ºC within ± 0.1 ºC. For
the 1000 ºC measurements, the sensitivity coefficient associated with the uncertainty in the reference
junction temperature is the ratio of those at 0 ºC and 1000 ºC, i.e.  0.407.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 60 OF 82
K7.11 The temperature gradients inside the furnace had been measured. At 1000 ºC deviations from non
uniformity of temperature in the region of measurement are within ± 1 ºC
K7.12 The compensation leads had been tested in the range 0 ºC to 40 ºC. Voltage differences between the
leads and the thermocouple wires were estimated to be less than ± 5 μV.
K7.13 The error due to inhomogeneity of the unknown thermocouple was determined during the calibration by
varying the immersion depth. Corrections are not practical for this effect therefore the error was
assumed to be zero within ± 0.3 ºC.
K7.14 The sequence of measurements is as follows: 1 First standard thermocouple
2 Unknown thermocouple
3 Second standard thermocouple
4 Unknown thermocouple
5 First standard thermocouple
The polarity is then reversed and the sequence is repeated. Four readings are thus obtained for all the
thermocouples. This sequence reduces the effects of drift in the thermal source and parasitic
thermocouple voltages. The results were as follows:
Thermocouple: First standard Unknown Second standard
Corrected voltages: + 10500 μV
+ 10503 μV
– 10503 μV
– 10504 μV
+ 36245 μV
+ 36248 μV
– 36248 μV
– 36251 μV
+ 10503 μV
+ 10503 μV
– 10505 μV
– 10505 μV
Absolute mean values: 10502.5 μV 36248 μV 10504.0 μV
Temperature of hot junctions: 1000.4 ºC 1000.6 ºC
Mean temperature of furnace: 1000.6 ºC
K7.15 The thermocouple output emf is corrected for the difference between the nominal temperature of 1000
ºC and the measured temperature of 1000.5 ºC. The reported thermocouple output is
5 . 1000
1000
36248× =
X
V µV = 36230 μV.
K7.16 In this example it is assumed that the procedure requires that the difference between the two standards
must not exceed 0.3 ºC. If this is the case then the measurement must be repeated and/or the reason
for the difference investigated.
K7.17 From the four readings on each thermocouple, one observation of the mean voltage of each thermocouple
was deduced. The mean voltages of the reference thermocouples are converted to temperature
observations by means of temperature/voltage relationships given in their calibration certificates. These
temperature values are highly correlated. By taking the mean they are combined into one observation of
the temperature of the furnace at the location of the test thermocouple. In a similar way one observation of
the voltage of the test thermocouple is extracted.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 61 OF 82
K7.18 In order to determine the random uncertainty associated with these measurements a Type A evaluation
had been carried out on a previous occasion. A series of ten measurements had been undertaken at the
same temperature of operation. Using Equation (7), this gave estimates of the standard deviations for
the temperature of the furnace, s
p
(t
S
) of 0.10 ºC and the voltage of the thermocouple to be calibrated,
s
p
(V
iX
), of 1.6 μV.
The resulting standard uncertainties were as follows:
From Equation (6): ( ) ( )
( )
1
1 . 0
= = =
n
t s
t s t u
S p
S p S
= 0.10 ºC,
and ( ) ( )
( )
1
6 . 1
= = =
n
V s
V s V u
iX p
iX p iX
= 1.6 μV.
The value of n = 1 is used to calculate the standard uncertainty because in the normal procedure only
one sequence of measurements is made at each temperature.
K7.19 Uncertainty budget  temperature of the furnace
Symbol Source of uncertainty
Value
±
Probability
distribution
Divisor c
i
u
i
(T)
ºC
ν
i
or
ν
eff
δt
S
Calibration of standard thermocouples
0.3 ºC Normal 2.0 1.0 0.150
∞
δt
D
Drift in standard thermocouples 0.3 ºC Rectangular √3 1.0 0.173 ∞
δV
iS1
Voltmeter calibration 2.0 μV Normal 2.0 0.077 0.077 ∞
δV
R
Switch contacts 2.0 μV Rectangular √3 0.077 0.089 ∞
δt
0S
Determination of reference point 0.1 ºC Rectangular √3 0.407 0.024 ∞
t
S
Repeatability 0.1 ºC Normal 1.0 1.0 0.10 9
δV
iS2
Voltmeter resolution 0.5 μV Rectangular √3 0.077 0.022
∞
δt
F
Furnace nonuniformity 1.0 ºC Rectangular √3 1.0 0.577 ∞
u
c
(T) Combined standard uncertainty
Normal 0.641 >500
K7.20 Uncertainty budget  emf of unknown thermocouple
Symbol Source of uncertainty
Value
±
Probability
distribution
Divisor c
i
u
i
(V)
μV
ν
i
or
ν
eff
Δt
X
Furnace temperature
0.641 ºC Normal 1.0 38.6 24.7 >500
δV
LX
Compensating leads
5.0 μV Rectangular √3 1.0 2.89 ∞
δV
iX1
Voltmeter calibration 2.0 μV Normal 2.0 1.0 1.0
∞
δV
R
Switch contacts 2.0 μV Rectangular √3 1.0 1.155 ∞
δt
0X
Determination of reference point 0.1 ºC Rectangular √3 26.1 1.48
∞
u(V
iX
) Repeatability 1.6 μV Normal 1.0 1.0 1.6 9
δV
iS2
Voltmeter resolution 0.5 μV Rectangular √3 1.0 0.29 ∞
h
TH
Inhomogeneity of thermocouple 0.3 ºC Rectangular √3 38.6 6.686 ∞
u
c
(V) Combined standard uncertainty Normal 25.9 >500
U Expanded uncertainty
Normal
(k = 2)
51.9 >500
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 62 OF 82
K7.21 Reported result
The type N thermocouple shows, at the temperature of 1000.0 °C and with its cold junction at a
temperature of 0 ºC, an emf of 36 230 μV ± 52 μV.
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2,
providing a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in
accordance with UKAS requirements.
K8 Calibration of a Digital Pressure Indicator (DPI) at a nominal pressure of 2 MPa using a reference
hydraulic dead weight tester
K8.1 The pressure was generated using a dead weight tester (DWT) the performance characteristics of
which had previously been determined. The indication was approached with increasing pressure to
account for the existence of possible hysteresis in the DPI. The indication of the unknown DPI is
obtained from:
( ) ( )
( ) ( ) ( )
( ) ( )
V R d a f
p
a f V
m
a
L
X
e I I gh
t a A
C g B g m
I + + + ÷ +
÷ + +
+ ÷ ÷ +
(
(
¸
(
¸


.

\

÷
=
¿
o µ µ
ì
o µ µ
µ
µ
20 1 1
1
0
where
I
X
= Indication of unknown DPI
m
L
= Mass of the component parts of the load, including the piston
ρ
a
= Density of ambient air
ρ
m
= Density of the mass, m, and can be significantly different for each load component.
g = The value of the local acceleration due to gravity
h = Height different between reference level of standard and reference level of generated
pressure
B
V
= Buoyancy volume of the reference piston – from calibration certificate.
p
f
= Density of hydraulic fluid
σ = Surface tension coefficient of hydraulic fluid
C = Circumference of reference piston
A
0
= Effective area at zero pressure of reference piston
a
p
= Distortion coefficient of reference piston (pressure dependant term)
λ = Temperature coefficient of piston and cylinder
δI
d
= The rounding of the value of one changing digit of the indication
I
R
= Repeatability of indication
e
V
= Error due to the piston not being perfectly vertical.
K8.2 The calibration certificate for the reference DWT gives the piston area and its uncertainty as:
A
0
= 80.6516 mm
2
± 0.0026 mm
2
. This results in a relative uncertainty in A
0
of 32.2 ppm.
K8.3 The calibration certificate for the reference DWT gives the distortion coefficient of the reference piston
as a
p
= 6.0 x 10
6
/MPa ± 0.5 x 10
6
/MPa.
K8.4 The drift limit in the effective area of the DWT, Ap
D
, based on results from previous calibrations, has
been set to ± 30 ppm.
K8.5 Mass uncertainties
K8.5.1 The mass of the piston is shown on the calibration certificate as 0.567 227 kg ± 0.000 010 kg.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 63 OF 82
K8.5.2 The drift of the piston mass, based on previous calibrations, has been set to 0.000 015 kg.
K8.5.3 The uncertainty of the mass set is shown on the calibration certificate for the three weights “A”, “B” and
“C” used to generate 2 MPa as:
A = 0.255 242 kg ± 0.000 010 kg
B = 7.402 137 kg ± 0.000 050 kg
C = 8.224 784 kg ± 0.000 050 kg
K8.5.4 The limits to the drift of the mass set have been set to be equal to the expanded uncertainty of its
calibration, i.e. 10 mg, 50 mg and 50 mg respectively.
K8.5.5 To obtain the combined mass uncertainty, u(m), in relative terms:
Relative uncertainty in mass u(m) =
(kg) C B A Piston of Mass
(mg) C B A Piston of y Uncertaint
+ + +
+ + +
ppm
=
784 224 . 8 137 402 . 7 242 255 . 0 227 567 . 0
50 50 10 10
+ + +
+ + +
= 7.30 ppm
Similarly, for drift this gives m
D
= 7.60 ppm.
K8.6 The uncertainty in the temperature of the piston, u
t
, coming from the analysis of the temperature indictor
used has been set as ± 0.5ºC. This will affect the pressure generated in proportion to the temperature
coefficient of the piston and cylinder combination. In this case a steel piston and cylinder has a
temperature coefficient of 23 ppm/ºC. This figure was obtained from the calibration certificate for the
DWT.
K8.7 A correction had been made for the value of the local acceleration due to gravity. This value had been
estimated from knowledge of the measurement location and took the Bouguer anomalies into account.
The expanded uncertainty associated with this estimate, U
g
, was ± 3 ppm (k = 2).
K8.8 No correction has been made for air buoyancy therefore, as all masses have an assumed density of
7800 kgm
3
, the uncertainty, a
b
, is estimated to be ± 13 ppm.
K8.9 The uncertainty relating to fluid head effects, f
h
, arises from the height different between the reference
level of the reference DWT and the generated pressure datum point. It is estimated as +2 mm and the
uncertainty in this estimate is ± 1 mm. No correction is made therefore a limit value of ± 3 mm has been
assigned for the uncertainty associated with fluid head effects. Assuming that the density of the oil used
is 917 kg/m
3
and the local value of g is 9.81 ms
 2
, then the uncertainty associated with the fluid head
effect is 917 x 9.81 x 0.003 = 27.0 Pa. In relative terms this corresponds to an uncertainty of ± 13.5 ppm
at 2 MPa.
K8.10 The uncertainty contribution, f
e,
from buoyancy volume, surface tension and fluid density effects has
been estimated as ± 2 ppm based on a relative uncertainty in each of ± 10%.
K8.11 The repeatability of the calibration process had been established from previous measurements using DPIs
of similar type and nominal range. This Type Aevaluation, based upon 10 measurements and using
Equation (7), yielded a relative standard deviation s(P
R
) of 16 ppm.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 64 OF 82
K8.12 An uncertainty arises due to the fact that the piston may not be perfectly vertical. If it were, then all of the
force would act on the area. Any departure from vertical will reduce the force  and therefore the pressure
 by the cosine of the angle. In this example, it is assumed that, after levelling, the piston is vertical to
within 0.15°. The effect in terms of generated pressure is proportional to the cosine of the angle from true
vertical. The cosine of 0.15°is 0.999 996 6. The maximum error is therefore 3.4 ppm of the generated
pressure.
NOTE
This effect always acts in one direction, i.e. the generated pressure will always be smaller than that obtained if the
piston were truly vertical. As this uncertainty is small compared with others in this particular calibration, it is
convenient to treat it as bilateral.
K8.13 The calibration of the unknown DPI was established from a single measurement, from which the DPI
indication was 2.000 12 MPa for an applied pressure of 2 MPa. As the conditions were the same as for the
previous evaluation of repeatability the standard uncertainty due to repeatability can be obtained from this
previous value of standard deviation, with n = 1, because only one reading is made for the actual
calibration. Therefore I
R
=
( )
1
R
P s
= 16 ppm.
K8.14 No correction can be made for the rounding, δI
d
, due to the resolution of the digital display of the DPI. The
least significant digit on the range being calibrated of this particular DPI changes in steps of 200 Pa and
there is therefore a possible rounding error of ± 100 Pa or, in relative terms, ±50 ppm. The probability
distribution is assumed to be rectangular.
K8.15 Uncertainty budget
Symbol Source of uncertainty
Value
±
Probability
distribution
Divisor c
i
u
i
(P
X
)
ppm
ν
i
or
ν
eff
A
0
Calibration of DWT (area)
32.2 ppm Normal 2.0 1.0 16.25 ∞
a
p
Calibration of DWT (distortion)
0.5 ppm/MPa Normal 2.0 2.0 0.500 ∞
Ap
D
Drift in area
30 ppm Rectangular √3 1.0 17.32 ∞
(u)m Calibration of total load 7.3 ppm Normal 2 1.0 3.65 ∞
m
D
Drift of total load 7.6 ppm Rectangular √3 1.0 4.38 ∞
u
t
Temperature of the piston 0.5 ºC Rectangular √3 23 6.64 ∞
a
b
Air buoyancy 13 ppm Rectangular √3 1.0 7.51 ∞
U
g
Local gravity determination 3.0 ppm Normal 2 1.0 1.50 ∞
f
h
Fluid head effects 13.5 ppm Rectangular √3 1.0 7.79 ∞
f
e
Other fluid effects 2.0 ppm Rectangular √3 1.0 1.15 ∞
u
L
Levelling effects 3.4 ppm Rectangular √3 1.0 1.96 ∞
δI
d
Digital rounding error 50 ppm Rectangular √3 1.0 28.9 ∞
I
R
Repeatability of indication 16 ppm Normal 1 1.0 16.0 9
u(I
X
)
Combined standard uncertainty
Normal 43.0 >350
U
Expanded uncertainty
Normal
(k = 2)
86.0 >350
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 65 OF 82
K8.16 The generated pressure was calculated from the mass of the piston (K8.5.1), that of the mass set
(K8.5.3) and the following quantities:
Temperature of Piston: 21 ºC
Buoyancy volume of piston: 3.107 m
3
Air density: 1.2 kg/m
3
Local gravity: 9.811812 ms
2
Oil surface tension coefficient: 0.0315 N/m
This results in an applied pressure of 2.000 806 MPa.
K8.17 Reported result
The pressure was applied in an increasing direction until it reached a final value of 2.000 806 MPa. The
indication of the digital pressure indicator was 2.000 84 MPa ± 86 ppm.
The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2,
providing a coverage probability of approximately 95%. The uncertainty evaluation has been carried out in
accordance with UKAS requirements.
NOTES
1 In this example, the uncertainty due to resolution, δI
d
, is larger than any other contribution and is assigned a rectangular
distribution. Nevertheless, the combined standard uncertainty is still Gaussian, due to the presence of the other uncertainties,
even though they are of smaller magnitude. This has been verified by Monte Carlo Simulation techniques.
2 The resolution uncertainty is based on the least significant digit of the DPI. However, this changes in steps of 2 digits for this
particular DPI, therefore the digital rounding error is 1 digit.
3 This uncertainty budget has been constructed in relative terms (ppm), as most of the errors that arise are proportional to the
generated pressure and it is the convention in this particular field of measurement to express uncertainties in this manner. If it
is required that the uncertainty is reported in absolute units, it can be calculated from the reported value and the relative
uncertainty. In this case, the expanded uncertainty in absolute terms is ± 0.000 17 MPa.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 66 OF 82
APPENDIX L
EXPRESSION OF UNCERTAINTY FOR A RANGE OF VALUES
L1
Introduction
L1.1 On occasions it is convenient to provide a statement of uncertainty that describes a range of values
rather than a single result.
L1.2 The GUM deals with expression of uncertainties for the reporting of a single value of a measurand, or
more than one parameter derived from the same set of data. In practice many measuring instruments
are calibrated at several points on a range and the use of an expression describing the uncertainty at
any of these points can be desirable.
L1.3 This Appendix therefore describes the situations when this can occur, explains how it can be dealt with
using the principles of this code of practice and provides an illustration of the process using a worked
example.
L2 Principles
L2.1 When measurements are made over a range of values and the corresponding sources of uncertainty are
examined it may be found that some are absolute in nature (i.e. they arise in a manner that is
independent of the value of the measurand) and some are relative in nature (i.e. they arise in a manner
that makes them proportional to the value of the measurand).
L2.2 It is possible, of course, to calculate a value for the expanded uncertainty for each reported value over
the range. This can give problems when reporting values near zero as a relative term may not be
appropriate and an absolute term has to be used. Conversely, when reporting values higher up the scale
it may be desirable to express the uncertainty in relative terms, as this is often how instrument
specifications are expressed.
L2.3 If the instrument being calibrated is subsequently to be used in a situation where a further analysis of
uncertainty is required, the user may also require to express these uncertainties in both absolute and
relative terms. However if the user has only been provided with a single value of uncertainty for each
reported value, it would not be possible to extract the absolute and relative parts from these single
values. Reporting uncertainties in both absolute and relative terms therefore provides more information
to the user than if a series of single values are quoted. Additionally, it is often more representative of the
way instrument specifications are expressed.
L2.4 The process of calculating a value of expanded uncertainty describing a range of values is identical to
that for single values except that the absolute and relative terms are identified as such and, in effect, a
separate uncertainty evaluation is carried out for each. These evaluations are carried out in the manner
already described in this publication.
L2.5 The results of these evaluations are then expressed as separate absolute and relative terms.
Traditionally this has been expressed in the form
± (U
REL
+ U
ABS
)
This linear addition of quantities is not in accordance with the principles embodied in the GUM, unless
there happens to be a high degree of correlation between the absolute and relative terms  which is not
usually the case. The two values should normally be reported separately with an appropriate statement
describing how they should be combined. A suggested statement is given in the example in L3.13.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 67 OF 82
L3 Example of uncertainty evaluation for a range of values
L3.1 In this example a 6½digit electronic multimeter is calibrated on its 1 V dc range using a multifunction
calibrator.
L3.2 The calibrations were carried out in both polarities at 0.1 V increments from zero to 1 V and additionally
at 1.5 V and 1.9 V. Only one measurement was carried out at each point and therefore reliance was
placed on a previous evaluation of repeatability using similar multimeters.
L3.3 No corrections were made for known errors of the calibrator as these were identified as being small
relative to other sources of uncertainty. The uncorrected errors are assumed to be zero with an
uncertainty obtained by analysis of information obtained from the calibration certificate for the calibrator.
L3.4 The indication of the multimeter under test, I
DVM
, can be described as follows:
I
DVM
= V
CAL
+ V
D
+ V
UE
+ V
TC
+ V
LIN
+ V
T
+ V
CM
+ δV
RES
+ I
R
, where
V
CAL
= Calibrated voltage setting of multifunction calibrator.
V
D
= Drift in voltage of multifunction calibrator since last calibration.
V
UE
= Uncorrected errors of multifunction calibrator.
V
TC
= Temperature coefficient of multifunction calibrator.
V
LIN
= Linearity and zero offset of multifunction calibrator.
V
T
= Thermoelectric voltages generated at junctions of connecting leads, calibrator and
multimeter.
V
CM
= Effects on voltmeter reading due to imperfect commonmode rejection characteristics of the
measurement system.
δV
RES
= Rounding errors due to the resolution of multimeter being calibrated.
I
R
= Repeatability of indication
L3.5 The calibration uncertainty was obtained from the certificate for the multifunction calibrator. This had a
value of ± 2.8 ppm as a relative uncertainty but there was an additional ± 0.5 μV in absolute units (k = 2).
L3.6 The manufacturer's 1year performance specification for the calibrator included the following effects:
V
D
, V
UE
, V
TC
These contributions were assumed to be relative in nature.
V
LIN
This contribution was assumed to be absolute in nature.
L3.7 The specification for the calibrator on the 1 V dc range was ± 8 ppm of reading ± 1 ppm of full scale. On
this particular multifunction calibrator the full scale value is twice the range value; therefore the absolute
term is ± 2 V x 10
6
= ± 2 μV. The performance of the calibrator had been verified by examining its
calibration data and history, using internal quality control checks and ensuring that it was used within the
temperature range and other conditions as specified by the manufacturer. A rectangular distribution was
assumed.
L3.8 The effects of thermoelectric voltages, V
T
, for the particular connecting leads used had been evaluated
on a previous occasion. Thermoelectric voltages are independent of the voltage setting are therefore an
absolute uncertainty contribution. A value of ± 1 μV was assigned, based on previous experiments with
the leads. The probability distribution was assumed to be rectangular.
L3.9
Effects due to commonmode signals, V
CM
, had also been the subject of a previous evaluation and a
value of ± 1 μV, with a rectangular distribution, was assigned. This contribution is absolute in nature, as
the commonmode voltage is unrelated to the measured voltage.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 68 OF 82
L3.10 No correction is made for the rounding due to the resolution V
RES
of the digital display of the multimeter.
The least significant digit on the range being calibrated corresponds to 1 μV and there is therefore a
possible rounding error δV
RES
of ± 0.5 μV. The probability distribution is assumed to be rectangular and
this term is absolute in nature.
L3.11 A previous evaluation had been carried out on the repeatability of the system using a similar voltmeter.
Ten measurements were carried out at zero voltage, 1 V and 1.9 V. Repeatability at the zero scale point
was found not to be significant compared with other absolute contributions. From Equation (7), this was
found to give a standard deviation s(V
i
) of 2.5 ppm. Then, from Equation (6);
( )
( )
1
5 . 2
= = =
n
V s
V s I
i
i p R
= 2.5 ppm.
L3.12 Uncertainty budget
Symbol Source of uncertainty
value
(relative)
± ppm
value
(absolute)
± μV
Probability
distribution
Divisor c
i
u
i
(V)
(relative)
ppm
u
i
(V)
(absolute)
μV
ν
i
or
ν
eff
V
CAL Calibration uncertainty
2.8 0.5 Normal 2.0 1 1.40 0.25 ∞
V
SPEC
Specification of calibrator 8.0 2.0 Rectangular √3 1 4.62 1.15 ∞
VT Thermoelectric voltages 1.0 Rectangular √3 1 0.58 ∞
V
CM
Common mode effects 1.0 Rectangular √3 1 0.58 ∞
δV
RES
Voltmeter resolution 0.5 Rectangular √3 1 0.29 ∞
I
R
Repeatability 2.5 Normal 1.0 1 2.5 9
u
c
(V)
Combined standard
uncertainty
Normal 5.42 1.46 >100
U Expanded uncertainty
Normal
(k = 2)
10.8 2.92 >100
L3.13 It is assumed that the results of this calibration will be presented in tabular form. After the results the
following statements regarding uncertainty can be given:
The expanded uncertainty for the above measurements is stated in two parts:
Relative uncertainty: ± 11 ppm
Absolute uncertainty: ± 2.9 μV
The reported twopart expanded uncertainty is in each case based on a standard uncertainty multiplied by a
coverage factor k = 2, providing a coverage probability of approximately 95%. The uncertainty evaluation has been
carried out in accordance with UKAS requirements. For each stated result the user may, if required, combine the
uncertainties shown by quadrature summation in either relative or absolute terms as appropriate.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 69 OF 82
APPENDIX M
ASSESSMENT OF COMPLIANCE WITH SPECIFICATION
M1 Introduction
M1.1 In many situations it will be necessary to make a statement in calibration certificates or test reports about
whether or not the reported result complies with a given specification.
M1.2 For calibration activities, this will often be the case for general purpose test and measurement
equipment. For measurement standards it is more likely that the measured value and expanded
uncertainty will be of more interest to the user, and specification compliance is less relevant.
M1.3 For testing activities it is very likely that the result will have to be compared with specified limits in order
to arrive at a conclusion relating to conformance or fitness for purpose.
M2 Theory
M2.1 It has been shown elsewhere in this document that an uncertainty associated with a measurement result
will usually take the form of a normal or Gaussian distribution centred around the reported value y of the
measurand Y.
M2.2 If the result y and the entirety of the distribution lie within the specified limits then it is clear that
compliance with the specification has been demonstrated. Conversely, If the result y and the entirety of
the distribution lie outside the specified limits then noncompliance with the specification has been
demonstrated.
M2.3 A “perfect” normal distribution has tails that extend to infinity in each direction, implying that there will
always be some doubt about whether or not compliance has or has not been demonstrated. In practice,
an uncertainty budget is always based on a finite number of contributions, thereby limiting the amount by
which the tails of the distribution extend away from the reported value.
M2.4 Nevertheless, it can be the case that one, or both, of the tails of the distribution significantly overlap one,
or both, of the specification limits. This means that consideration has to be given to the area of the
distribution that is contained within the limits when assessing compliance with the specification.
M2.5 Furthermore, the implication of this is that compliance – or noncompliance – can only be stated in
conjunction with an associated confidence level. This is because there will always be a possibility of one,
or both, of the tails of the distribution overlapping the limits.
M2.6 This concept is illustrated in Figure 7 overleaf.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 70 OF 82
Figure 7
M2.7
Lower specification limit Result Upper specification limit
In this example, it can be seen that the majority of the distribution lies
within the specification limits, but a significant proportion lies above the
upper limit. If, for example, 95% of the area of the distribution is within the
specification limits, and 5% of it is outside them, then compliance has
been demonstrated at a confidence level of 95%. Similar reasoning
applies regarding noncompliance with the specification if the result were
to be outside the limits.
M2.8 As an expanded uncertainty is normally expressed for a coverage probability of approximately 95%, it is
generally accepted practice that statements regarding compliance will relate to the same level of
confidence.
M2.9 If this is the case for a given situation, then comparison of the expanded uncertainty using a coverage
factor k = 2 with the specification limit is unduly pessimistic. It will yield a confidence level of 97.5% or
greater. This is because only one tail of the distribution will usually be the subject of comparison with the
limit. If, as should be the case, the uncertainty is small compared with the specification, the probability
contained within the other tail will already be within the specification limits.
M2.10 A new coverage factor, k
s
, should therefore be used for the purpose of comparison with a specification
limit. Assuming a normal distribution, the value of k
s
required to achieve at least 95% confidence is 1.64.
M2.11
NOTE
It is common practice for a compliance decision to be made by simply comparing the expanded uncertainty, with a
coverage probability of approximately 95%, with the specification. This will always yield a “safe” decision, however the
method described above is preferred as it takes into account the actual amount of probability that breaches the
specification limit.
M2.12 It may be the case that compliance, or noncompliance, with a specification cannot be demonstrated at
95% confidence, as described above. One solution is to reduce the uncertainty, possibly by applying
corrections to the result, using more accurate equipment or by taking the mean of a larger amount of
readings. If this is not practical or desirable, then it may be possible to evaluate compliance, or non
compliance, at a different level of confidence. Such an approach should, of course, be taken with the
agreement and understanding of the customer.
M2.13 This procedure may also be used in cases where a customer has requested a compliance statement for
a confidence level other than 95%.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 71 OF 82
M2.14 Two pieces of information are needed to deduce the confidence level at which compliance can be
stated:
The combined standard uncertainty u
c
(y)
The difference between the specification limit and the result, L
S
 y
M2.15 The probability of compliance when the result lies within the specification, or that of noncompliance
when it lies outside the specification, can be obtained from the following table:
M2.16
Probability of
compliance or
noncompliance
( ) y u
y L
c
S
÷
Probability of
compliance or
noncompliance
( ) y u
y L
c
S
÷
Probability of
compliance or
noncompliance
( ) y u
y L
c
S
÷
99.9 3.29 91 1.34 80 0.84
99.73 2.78 90 1.28 79 0.81
99 2.32 89 1.23 78 0.77
98 2.05 88 1.17 77 0.74
97 1.88 87 1.13 76 0.71
96 1.75 86 1.08 75 0.68
95.45 1.69 85 1.04 74 0.64
95 1.64 84 1.00 73 0.61
94 1.56 83 0.95 72 0.58
93 1.48 82 0.92 71 0.55
92 1.41 81 0.88 70 0.52
M2.17
Normally
( ) y u
y L
c
S
÷
will not be an integer and it will be necessary to interpolate between the values given in
the table. Linear interpolation will suffice for
( ) y u
y L
c
S
÷
<2; higherorder interpolation should be used
otherwise. Alternatively, the next lower value may be used.
M2.18 It should be noted that this procedure is only valid when the uncertainty breaches one of the
specification limits and for this reason the uncertainty should be sufficiently small that an insignificant
portion of the distribution approaches the other limit.
M2.19 Furthermore, as the result approaches either limit there will come a point at which no reasonable
decision can be made regarding compliance or noncompliance with the specification. In the extreme, if
the result coincided exactly with one of the limits, there would always be 50% confidence in the
decision, regardless of the magnitude of the uncertainty. For this reason, and by general convention, the
table above is limited to a confidence probability of 70% and above.
M2.20
Example
A measurement yields a result y of 0.80 units with a combined standard uncertainty u
c
(y) of 0.15 units.
The specification is ± 1.00 unit. At what confidence level can compliance with the specification be
made?
( ) y u
y L
c
S
÷
=
15 . 0
80 . 0 00 . 1 ÷
= 1.33. The next lower value in the table is 1.28, therefore it has been
demonstrated that the specification is met for at least 90% confidence.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 72 OF 82
M2.21 It may be the case that a result cannot be demonstrated to comply with a specification for a given
confidence level, but only a statement of the likelihood of compliance is required. An example of how
this may be reported is given below:
M2.22
The measured result is below (above) the specification limit by a margin less than the measurement uncertainty; it
is therefore not possible to state compliance (noncompliance) based on the stated coverage probability. However
the result indicates that compliance (noncompliance) is more probable than noncompliance (compliance) with the
specification limit.
M2.23 In cases such as those described above it is essential that the client is made aware of the situation
because the enduser is taking some of the risk that the item may not meet the specification. It should
also be noted that this “shared risk” approach may be superseded by legal requirements; for example, it
is not permitted in some areas of legal metrology.
M3 Specification limits – further considerations
In Section M2 the specification limits have been treated as absolute, analogous to a rectangular
probability distribution. This may not always be the case.
M3.1 There are situations where the specification is characterised by a normal distribution. It is stated in the
GUM that, when a specification is quoted for a given coverage probability, then a normal distribution can
be assumed. Some manufacturers state confidence levels for their specifications.
M3.2 If both the uncertainty U and the specification L are stated at the same coverage probability, then the
specification is met at that coverage probability when y <
2 2
U L
S
÷ and is failed when y >
2 2
U L
S
+ ,
where
y = reported result
U= expanded uncertainty
L
S
= specification limit
A worked example is shown below.
M3.3 Example
A digital multimeter is calibrated with an applied voltage of 10 .000 000 V dc. The expanded uncertainty,
U, is ± 3 ppm (approximately 95% confidence, k = 2) and the reading is + 7 ppm from the nominal value.
The manufacturer’s specification for this reading is stated as 10 ppm at 99% confidence. Is the
specification met?
The comparison with specification has to be carried out with both the specification and the uncertainty at
the same coverage probability. The specification at the 95.45% coverage probability, L
95.45
, can be
obtained from
99
45 . 95
99 45 . 95
k
k
L L · =
From Paragraph 3.47, k
95.45
= 2 and k
99
= 2.58, therefore
58 . 2
0 . 2
10
45 . 95
· = L = 7.75 ppm.
2
45 . 95
2
45 . 95
U L ÷ =
2 2
3 75 . 7 ÷ = 7.15 ppm. As 7 < 7.15, then compliance with the specification has been
demonstrated, taking the measurement uncertainty into account.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 73 OF 82
M4 Reporting compliance with specification
M4.1 If compliance or noncompliance with a specification is clearly demonstrated for a given confidence level
then a statement to this effect can be made in calibration certificates and test reports. However care
must be taken to ensure that there is no implication that parameters that have not been measured also
comply with a specification. For this reason a broad statement such as "the equipment complies with its
specification" should not be made. A suggested statement of compliance is as follows:
The equipment complies with the stated specification at the measured points for the stated confidence level, due
allowance having been made for the uncertainty of the measurements.
This statement can be modified as necessary where noncompliance with a specification is to be
reported.
M4.2 When making compliance or noncompliance statements, the specification and any relevant clauses
within it should be unambiguously identified in the calibration certificate or test report.
M4.3 There may be cases where the uncertainty attainable for a given test or calibration is larger than the
specification for the item under consideration. Such situations should be subject to the contract
arrangements between the laboratory and its customer. Astatement regarding compliance should not be
given under these circumstances but a note should be included indicating those uncertainties associated
with the measured values that are greater than the required specifications.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 74 OF 82
APPENDIX N
UNCERTAINTIES FOR TEST RESULTS
N1 Introduction
N1.1 ISO/IEC 17025:2005 requires that “testing laboratories shall have and apply procedures for estimating
uncertainty of measurement”.
N1.2 It is recognised that the present state of development and application of uncertainties in testing activities
is not as comprehensive as in the calibration fields, to which much of this document is addressed. It is
therefore accepted that the implementation of ISO/IEC 17025:2005 criteria on this subject will take place
at an appropriate pace, which may differ from one field to another. However laboratories should be able
to satisfy requests from clients, or requirements of specifications, to provide statements of uncertainty.
N1.3 Testing laboratories should therefore have a defined policy covering the evaluation and reporting of the
uncertainties associated with the tests performed. The laboratory should use documented procedures
for the evaluation, treatment and reporting of the uncertainty.
N1.4 Some tests are qualitative in nature, i.e., they do not yield a numeric result. Therefore there can be no
meaning in reporting uncertainties directly associated with the test result. Nevertheless, there will be
uncertainties associated with the underlying test conditions and these should be subject to the same
type of evaluation as is required for quantitative test results.
N1.5 The methodology for estimation of uncertainty in testing is no different from that in calibration and
therefore the procedures described in this document apply equally to testing results.
N2 Objectives
N2.1 The objective of a measurement is to determine the value of the measurand, ie the specific quantity
subject to measurement. When applied to testing, the general term measurand may cover many
different quantities, for example:
the electrical breakdown characteristics of an insulating material;
the strength of a material;
the concentration of an analyte;
the level of emissions of electromagnetic radiation from an appliance;
the quantity of microorganisms in a food sample;
the susceptibility of an appliance to electric or magnetic fields;
the quantity of asbestos particles in a sample of air.
N2.2 A measurement begins with an appropriate specification of the measurand, the generic method of
measurement and the specific detailed measurement procedure. Knowledge of the influence quantities
involved for a given procedure is important so that the sources of uncertainty can be identified.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 75 OF 82
N3 Sources of uncertainty
N3.1 There are many possible sources of uncertainty. As these will depend on the nature of the tests involved,
it is not possible to give detailed guidance here. However the following general points will apply to many
areas of testing:
(a) Incomplete definition of the test  the requirement may not be clearly described, e.g. the temperature of a
test may be given as 'room temperature'.
(b) Imperfect realisation of the test procedure; even when the test conditions are clearly defined it may not
be possible to produce the theoretical conditions in practice due to unavoidable imperfections in the
materials or systems used.
(c) Sampling  the sample may not be fully representative. In some disciplines, such as microbiological
testing, it can be very difficult to obtain a representative sample.
(d) Inadequate knowledge of the effects of environmental conditions on the measurement process, or
imperfect measurement of environmental conditions.
(e) Personal bias and human factors; for example:
 Reading of scales on analogue indicating instruments.
 Judgement of colour.
 Reaction time, e.g. when using a stopwatch.
 Instrument resolution or discrimination threshold, or errors in graduation of a scale.
(f) Values assigned to measuring equipment and reference materials.
(g) Changes in the characteristics or performance of measuring equipment or reference materials since the
last calibration.
(h) Values of constants and other parameters used in data evaluation.
(i) Approximations and assumptions incorporated in the measurement method and procedure.
(j) Variations in repeated observations made under similar but not identical conditions  such random
effects may be caused by, for example, electrical noise in measuring instruments, shortterm fluctuations
in the local environment, e.g. temperature, humidity and air pressure, variability in the performance of
the person carrying out the test and variability in the homogeneity of the sample itself.
N3.2 These sources are not necessarily independent and, in addition, unrecognised systematic effects may
exist that cannot be taken into account but contribute to error. This is one reason that participation in
interlaboratory comparisons, participation in proficiency testing schemes and internal crosschecking of
results by different means are encouraged.
N3.3
Information on some of the sources of these errors can be obtained from:
(a) Data in calibration certificates  this enables corrections to be made and uncertainties to be assigned.
(b) Previous measurement data  for example, history graphs can be constructed and can yield useful
information about changes with time.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 76 OF 82
(c) Experience with or general knowledge about the behaviour and properties of similar materials and
equipment.
(d) Accepted values of constants associated with materials and quantities.
(e) Manufacturers' specifications.
(f) All other relevant information.
These are all referred to as Type B evaluations because the values were not obtained by statistical
means. However the influence of random effects is often evaluated by the use of statistics; if this is the
case then the evaluation is designated Type A.
N3.4 Definitions are given in paragraph 3.10 for Type A evaluations and in paragraph 3.11 for Type B
evaluations. Further detail on the means of evaluation is given in Sections 4 and 5.
N3.5 It is recognised that in certain areas of testing it may be known that a significant contribution to
uncertainty exists but that the nature of the test precludes a rigorous evaluation of this contribution. In
such cases, ISO/IEC 17025:2005 requires that a reasonable estimation be made and that the form of
the reporting does not give an incorrect impression of the uncertainty.
N3.6 In some fields of testing it may be the case that the contribution of measuring instruments to the overall
uncertainty can be demonstrated to be insignificant when compared with the repeatability of the process.
Nevertheless, such instruments have to be shown to comply with the relevant specifications, normally by
calibration.
N3.7 Some analysis processes appear at first sight to be quite complex, for example there may be various
stages of weighing, dilutions and processing before results are obtained. However it will sometimes be
the case that the procedure requires standard reference materials to be subject to the same process, the
result being the difference between the readings for the analyte and the reference material. In such
cases, much of the process can be considered to be negatively correlated and the uncertainty of
measurement can be evaluated from the resolution and repeatability of the process; matrix effects may
also have to be considered.
N4 Process
N4.1 The process of assigning a value of uncertainty to a measurement result is summarised below:
(a) Identify all sources of error that are likely to have a significant effect, and their relationship with the
measurand.
(b) Assign values to these using information such as described in N3.3, or in the case of Type A evaluations,
calculate the standard deviation using Equations (5) and (6).
(c) Consider each uncertainty component and decide whether any are interrelated and whether a dominant
component exists (see D3 and Appendix C respectively).
(d) Add any interdependent components algebraically (i.e., account for whether they act together or cancel
each other) and derive a net value.
(e) Express each uncertainty value as the equivalent of a standard deviation (see paragraph 3.24), taking
into account any sensitivity coefficients (see paragraphs 3.28 to 3.35).
(f) Take the independent components and the values of any derived net components and, in the absence of
a dominant component, combine them by taking the square root of the sum of the squares. This gives
the combined standard uncertainty (Equation (1)).
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 77 OF 82
(g) Multiply the combined standard uncertainty by a coverage factor k, selected on the basis of the coverage
probability required, to provide the expanded uncertainty U (see paragraphs 3.42 to 3.44).
(h) Report the result and, if required, the expanded uncertainty, coverage factor and coverage probability in
accordance with Section 6.
N4.2 If one uncertainty contribution is significantly larger than the others then modifications may be required
to this procedure. In the case of a dominant component derived from Type B evaluation, see Appendix
C. If the nonrepeatability of the system is significant, and its effects are evaluated by using a Type A
analysis, it may be necessary to use the procedure in Appendix B.
N4.3 Further information regarding uncertainty evaluation for testing activities can be obtained from specialist
publications that address particular fields of testing, such as are described in References [7] and [8].
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 78 OF 82
APPENDIX P
ELECTRONIC DATA PROCESSING
P1 Introduction
P1.1 Due to the nature and quantity of the calculations involved it is inevitable that some form of electronic
processing will be involved in these calculations. This Appendix gives brief details of precautions that
may be necessary under these circumstances. Mention is also made of other techniques of uncertainty
evaluation that electronic data processing has made practical.
P2 Use of calculators
P2.1 Most scientific calculators are easily capable of all the calculations required for the evaluation of
measurement uncertainty. It is recommended that readers of this document gain fami liarity with the
functions involved by practicing the calculations presented in Appendix K  this can give rise to a better
understanding of the process as well as giving the users confidence in their own abilities.
P2.2 It is also recommended that, where possible, intermediate results are stored in the calculator memory for
use later, or are written down to a reasonable amount of significant figures, in order to prevent the
cumulative effects of rounding errors having a significant effect on the result. Most scientific calculators
work with sufficient accuracy so that they do not in themselves introduce any significant errors  with one
notable exception (see Section P4).
P3 Use of spreadsheets
P3.1 The widespread use of personal computers has made repetitive calculations a much easier process than
in the past. It is possible to construct a spreadsheet using the various equations presented in this
document in order to perform the uncertainty calculations. The time spent constructing the spreadsheet
can easily be recovered by the subsequent ease of producing, and amending, uncertainty budgets.
P3.2 It is recommended that a sample of results produced by a spreadsheet programme is checked manually
using a scientifi c calculator to ensure that correct results are being generated.
P4 Calculation of standard deviations
P4.1 Many scientific calculators include statistical functions for calculation of standard deviations in
accordance with Equation 5. The calculator key associated with this function will usually be marked σ
n1
or, sometimes, s.
P4.2 It will often be the case that this particular function is not capable of evaluating small values of standard
deviation correctly. The following data set is presented as an example:
1000.025 1000.015 1000.019 1000.021
P4.3 The value of s that is obtained from this data, using Equation 5, should be 0.004163. Most calculators
will either:
(a) Display an error message, or
(b) Display a value of zero, or
(c) Display an incorrect nonzero value.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 79 OF 82
P4.4 A solution to this problem is to use only the few least significant digits in the set of data. So, for the data
above, the numbers 0.025, 0.015, 0.019 and 0.021 can be entered, yielding the correct value of s(q
j
):
0.00416.
NOTE
This could equally have been evaluated using just the last two digits, i.e. 25, 15, 19 and 21. However it is useful to
include a decimal point, as this will then appear in the correct place in the results thereby minimising the likelihood
of errors being made.
P4.5 It is possible that electronic spreadsheets will also suffer from similar errors therefore suitable checks
should be devised to evaluate any such effects.
P5 Other techniques
P5.1 The use of high speed data processing has opened new avenues for the analysis of uncertainty. One
such technique is known as Monte Carlo simulation.
P5.2 In a Monte Carlo simulation the mathematical model of the underlying system is run over and over
again, each time using a different set of random numbers representing the input variables. Each of these
sets of random numbers combines via the model to represent a different outcome. Each run of the
model is called a simulation, or trial, and at the end of each trial the outcome of the process is recorded.
Each different outcome arises, through the measurement model, corresponding to a particular set of
random numbers being applied to it. If the model is a good representation of the realworld system, then,
by running a large enough number of trials (each with a different set of random numbers), the whole
range of possible outputs can be produced; these form the distribution of the output.
P5.3 Sampling techniques, such as Monte Carlo simulation, provide an alternative approach to uncertainty
evaluation in which the propagation of uncertainties is undertaken numerically rather than analytically.
Such techniques are useful for validating the results returned by the application of the GUM, as well as
i n circumstances where the assumptions made by the GUM do not apply. In fact, these techniques are
able to provide much richer information, by propagating the distributions (rather than just the
uncertainties) of the inputs x
i
through the measurement model f to provide the distribution of the output
y. From the output distribution confidence intervals can be produced, as can other statistical information.
P5.4 Another method of uncertainty analysis is by the use of Bayesian statistics; this is based on the concept
of degree of belief. Bayes’ Theorem gives the ’posterior’ odds on the correctness of a belief (given the
new evidence that has just been observed), making use of the ’prior’ odds that the belief is correct, i.e.
an estimate of the plausibility of the belief before one has the new data. A concept of likelihood ratio
allows the ‘prior’ odds to be adjusted according to the each piece of new evidence. Another iteration is
then performed using the next piece of evidence, and so on until all the evidence is used and a
probability associated with the belief is obtained.
P5.5 Bayes’ Theorem is particularly useful for analysis of qualitative data, where the conventional GUM
methodology is not appropriate. It can also be applied to the expression of professional opinions that
may appear in test reports.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 80 OF 82
APPENDIX Q
SYMBOLS
The symbols used are taken mainly from the GUM. The meanings have also been described further in the text,
usually where they first occur, but are summarised here for convenience of reference.
a
i
Estimated semirange of uncorrelated systematic component of uncertainty, probabil ity distributions
unknown, where i = 1 ..... N.
a
d
A systematic component of uncertainty that so dominates other contributions to uncertainty in
magnitude that special consideration has to be given to its presence in calculating the expanded
uncertainty.
c
i
Sensitivity coefficient used to multiply the value x
i
of an input quantity X
i
to express it in terms of the
measurand Y.
f Functional relationship between the measurand Y and the input quantities X
i
on which Y depends,
and between output estimate y and input estimates x
i
upon which y depends.
∂f/∂x
i
Partial derivative with respect to input quantity X
i
of the functional relationship f between the
measurand and the input quantities.
k Coverage factor (general).
k
p
Coverage factor used to calculate an expanded uncertainty U
p
for a specified coverage probability
p.
k
s
Coverage factor chosen for the purpose of comparison with a specification limit.
m Number of readings or observations that are used for the evaluation of s(q
j
), if different from n.
n Number of readings or observations that contribute to a mean value.
N Number of input estimates x
i
on which the value of the measurand depends.
q
j
jth repeated observation of randomly varying quantity q.
See Note 1 below.
q
Arithmetic mean or average of n repeated observations of randomly varying quantity q.
p Coverage probability or level of confidence expressed in percentage terms or in the range zero to
one.
σ The standard deviation of a population of data using all the samples in that population.
s(q
j
) Estimate of the standard deviation σ of the population of values of a random variable q based on a
limited sample of results from that population.
See Note 1 overleaf.
( ) q s
Experimental standard deviation of the mean value q .
t
p
(ν
eff
)
Student tfactor for ν
eff
degrees of freedom corresponding to a given coverage probability p.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 81 OF 82
u(x
i
)
Standard uncertainty of input estimate x
i
.
u
c
(y)
Combined standard uncertainty of output estimate y.
U Expanded uncertainty of output estimate y that describes the measurand as an interval Y = y ± U,
with a high coverage probability.
U
p
Expanded uncertainty of output estimate y that describes the measurand as an interval Y = y ± U
p
,
with a specified coverage probability p.
ν Degrees of freedom; in general, the number of terms in a sum minus the number of constraints on
the terms of the sum.
ν
i
Degrees of freedom of standard uncertainty u(x
i
) of input estimate x
i
.
ν
eff
Effective degrees of freedom of u
c
(y) used to obtain t
p
(ν
eff
)
NOTE 1 The GUM uses the symbols q
k
and s(q
k
) where q
j
and s(q
j
) are used here. M3003 uses the subscript j instead of k in order to
avoid any possible confusion with the coverage factor k.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN
Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 82 OF 82
APPENDIX R
REFERENCES
1 BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, OIML. Guide to the Expression of Uncertainty in
Measurement. International Organisation for Standardization, Geneva, Switzerland. ISBN 9267
101889, First Edition 1993.
BSI Equivalent:
BSI PD 6461: 1995, Vocabulary of metrology, Part 3. Guide to the Expression of Uncertainty in
Measurement. BSI ISBN 0 580 23482 7
2 European cooperation for Accreditation, EA4/02 Expression of the Uncertainty of Measurement in
Calibration, December 1999
3 HARRIS, I.A and WARNER, F.L. Reexamination of mismatch uncertainty when measuring
microwave power and attenuation. IEE Proceedings, Vol 128 Pt H No 1, February 1981.
4 DIETRICH, C.F. Uncertainty, Calibration and Probability: The Statistics of Scientific and Industrial
Measurement Uncertainty, Institute of Physics, January 1991.
5 National Physical Laboratory, ANAMET Connector Guide, Edition 2.
6 United Kingdom Accreditation Service, Calibration of Weighing Machines, LAB14 Edition 4,
November 2006.
7 United Kingdom Accreditation Service, The Expression of Uncertainty in EMC Testing, LAB34
Edition1, August 2002.
8 EURACHEM/CITAC Guide: Quantifying Uncertainty in Analytical Measurement, QUAM:2000.1,
Second Edition, 2000. ISBN 0 948926 15 5
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT
M3003  EDITION 2  JANUARY 2007
1
1.1
INTRODUCTION
The general requirements that testing and calibration laboratories have to meet if they wish to demonstrate that they operate to a quality system, are technically competent and are able to generate technically valid results are contained within ISO/IEC 17025:2005. This international standard forms the basis for international laboratory accreditation and in cases of differences in interpretation remains the authoritative document at all times. M3003 is not intended as a prescriptive document, and does not set out to introduce additional requirements to those in ISO/IEC 17025:2005 but to provide amplification and guidance on the current requirements within the international standard. The purpose of these guidelines is to provide policy on the evaluation and reporting of measurement uncertainty for testing and calibration laboratories. Related topics, such as evaluation of compliance with specifications, are also included. A number of worked examples are included in order to illustrate how practical implementation of the principles involved can be achieved. The guidance in this document is based on information in the Guide to the Expression of [1] Uncertainty in Measurement , hereinafter referred to as the GUM. M3003 is consistent with the GUM both in methodology and terminology. It does not, however, preclude the use of other methods of uncertainty evaluation that may be more appropriate to a specific discipline. For example, the use of Bayesian statistics is becoming recognised as being particularly useful in certain areas of testing. M3003 is aimed both at the beginner and at those more experienced in the subject of measurement uncertainty. In order to address the needs of an audience with a wide spectrum of experience, the subject is introduced in relatively straightforward terms and gives details of the basic concepts involved. Crossreferences are made to a number of Appendices, where more detailed information is presented for those who wish to obtain a deeper understanding of the subject. Edition 2 of M3003 is a complete revision of the previous issue and it is impractical to list all the changes in detail. Some of the more notable changes or additions are as follows: An overview of the basic concepts relating to uncertainty evaluation is given in Section 2 to introduce these ideas to those new to the subject. This is then expanded upon in Section 3, which gives more formal detail. ISO/IEC 17025:2005 criteria, which were not in place when Edition 1 was published, have been considered. A section on derivation of the measurement model has been included. Concepts are accompanied by simple worked examples as they are introduced. A number of diagrams illustrating the concepts have been included. Further detail has been included regarding dominant contributions. The subject of compliance with specification has been explored in more detail.
1.2
1.3
1.4
1.5
1.5.1
1.5.2 1.5.3 1.5.4 1.5.5 1.5.6 1.5.7
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 2 OF 82
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT
M3003  EDITION 2  JANUARY 2007
1.5.8 1.5.9
The worked examples have been reviewed and minor amendments have been made as necessary. A new example, relating to pressure calibration using a deadweight tester, has been added, together with a section describing the sources of uncertainty for these measurements. Hyperlinks have been included between various sections of the document. Analytical concepts such as Monte Carlo simulation and Bayesian statistics have been introduced. No further changes of any significance will be made to M3003 Edition 2 during its lifetime. However, minor text modifications of an editorial nature may be made if the need is identified. Any such changes will be listed below. Date 2 January 2007 8 January 2007 30 January 2007 22 February 2008 Details of amendment First issue of M3003 Edition 2 Amendment A1: Minor text corrections made; paragraph numbers in Appendix M corrected; crossreference on Page 16 corrected. Amendment A2: Probability distribution in second line of uncertainty budget K8.15 corrected; reference to LAB14 on page 82 updated. Amendment A3: Sensitivity coefficient CS in K7.6 corrected from 0.200 ºC/μ V to CS = 0.189 ºC/μ Resolution in K8.14 corrected from 20 Pa to 200 Pa. V. Repeated line removed in K8.1.
1.5.10 1.5.11 1.6
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 3 OF 82
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT
M3003  EDITION 2  JANUARY 2007
2
2.1
OVERVIEW
In many aspects of everyday life, we are accustomed to the doubt that arises when estimating how large or small things are. For example, if somebody asks, “what do you think the temperature of this room is?” we might say, “it is about 23 degrees Celsius”. The use of the word “about” implies that we know the room is not exactly 23 degrees, but is somewhere near it. In other words, we recognise that there is some doubt about the value of the temperature that we have estimated. We could, of course, be a bit more specific. We could say, “it is 23 degrees Celsius give or take a couple of degrees”. The term “give or take” implies that there is still doubt about the estimate, but now we are assigning limits to the extent of the doubt. We have given some quantitative information about the doubt, or uncertainty, of our estimate. It is also quite reasonable to assume that we may be more sure that our estimate is within, say, 5 degrees of the “true” room temperature than we are that it is within 2 degrees. The larger the uncertainty we assign, the more confident we are that it encompasses the “true” value. Hence, for a given situation, the uncertainty is related to the level of confidence. So far, our estimate of the room temperature has been based on a subjective evaluation. This is not entirely a guess, as we may have experience of exposure to similar and known environments. However, in order to make a more objective measurement it is necessary to make use of a measuring instrument of some kind; in this case we can use a thermometer. Even if we use a measuring instrument, there will still be some doubt, or uncertainty, about the result. For example we could ask: “Is the thermometer accurate?” “How well can I read it?” “Is the reading changing?” “I am holding the thermometer in my hand. Am I warming it up?” “The relative humidity in the room can vary considerably. Will this affect my results?” “Does it matter where in the room I take the measurement?” All these factors, and possibly others, may contribute to the uncertainty of our measurement of the room temperature.
2.2
2.3
2.4
2.5
2.6
In order to quantify the uncertainty of the room temperature measurement we will therefore have to consider all the factors that could influence the result. We will have to make estimates of the possible variations associated with these influences. Let us consider the questions posed above.
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 4 OF 82
Measuring instruments change their characteristics as time goes by. This leads to the concept of traceability of measurements. It is therefore important to evaluate the likely change since the instrument was last calibrated.35 °C and 23.4 °C therefore means that the underlying value is somewhere between 23. A reading of 23.ukas.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 2. otherwise the rounded reading would be 23.1 Is the thermometer accurate? In order to find out. If it is a liquidinglass thermometer.3 There are other possible influences relating to the thermometer accuracy.1 °C.2 In terms of the thermometer accuracy. Evaluations made using these methods yield the uncertainty due to secular stability. suppose we have a traceable calibration.8 2.45 °C. in turn. ISO/IEC17025:2005 requires that measurements are traceable to SI units. What does this tell us about its indication error at 23 °C? In such cases we will have to make an estimate of the error. the past data may not indicate a reliable trend. and so on. This is also known as “drift”. the finite number of digits in the display will define the limit. whereby measurements at all levels can be traced back to agreed references.4 °C.8. we need a traceable calibration.1 How well can I read it? There will inevitably be a limit to which we can resolve the reading we observe on the thermometer. Middlesex. Similarly. which maintains measurement standards that are directly related to SI units.1 °C resolution of the display has caused a rounding error somewhere between –0. In other words. it will be necessary to compare it with a thermometer whose accuracy is better known. or changes with time. In other words.8.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. suppose the last digit of a digital thermometer can change in steps of 0. a traceable calibration is not the end of the story. 2. and apply a correction to the reading. this is often referred to as the imported uncertainty. will have to be compared with an even better characterised one.7. This can be estimated from examination of changes that occurred in the past. This is usually achieved by an unbroken chain of comparisons to a national metrology institute. They “drift”. perhaps using interpolation between points where calibration data is available. When used in a subsequent evaluation of uncertainty. of course.7. In the case of a reading of 23.05 °C and +0.7. however. based on past results. such as the manufacturer’s specification. This thermometer. What does this mean in terms of uncertainty? The reading is a rounded representation of an infinite continuum of underlying values that the thermometer would indicate if it had more digits available. as the calibrating laboratory will assign a calibration uncertainty to the reported values. If the instrument has a reliable history it may be possible to predict what the reading error will be at a given time in the future. In most cases. It may then be necessary to use other information. 2. TW13 4UN Website: www. This prediction will not be perfect and therefore an uncertainty on the corrected value will be present. the underlying value cannot be more than 23. For example. This. to evaluate the additional uncertainty that arises when the reading is not directly at a point that has been calibrated.35 °C. but only at 15 °C. Feltham.5 °C.05 °C. 2. 2147 High Street. UKAS Copyright exists in all UKAS publications 2.45 °C.2 PAGE 5 OF 82 . This calibration itself will provide a source of uncertainty. This is not always possible as it depends on the measured data being such that accurate interpolation is practical.7 2. In other cases.4 °C. The reading happens to be 23. is why regular recalibration is necessary. For example. If it is a thermometer with a digital readout. 20 °C and 25 °C. this means that the underlying value cannot be less than 23. otherwise the rounded reading would be 23. and a limit value may have to be assigned for the likely change since the last calibration.3 °C. of the instrument. the 0. As we have no way of knowing where in this range United Kingdom Accreditation Service. this limit will often be imposed by our ability to interpolate between the scale graduations.
however it is more correctly the numeric rounding caused by finite resolution.7 United Kingdom Accreditation Service.456. or arithmetic mean.9. There will be an unknown error – and therefore an uncertainty – represented by the difference from our calculated mean value and the underlying “true” mean value. NOTE In earlier textbooks on the subjects of uncertainty or statistics. TW13 4UN Website: www. Feltham. it applies every time a number is recorded.9. This uncertainty is referred to as the experimental standard deviation of the mean.5 2.6 It is often convenient to regard the calculation of the standard deviation of the mean as a twostage process. Up until now. we will probably take an average of several measurements in order to obtain a more realistic reading. we have looked for evidence. So which is “correct”? In practice. 2. this may be referred to as the standard error of the mean. nonrepeatability) of our measurements. Is the reading changing? Yes.9 2.9.3 2.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 the underlying value is. UKAS Copyright exists in all UKAS publications PAGE 6 OF 82 . For the sake of brevity.9. we have to assume the rounding error is zero with limits of ± 0. of course. The effects of variation between readings cannot be evaluated like this. or mean. such as the way we are holding the thermometer. of a number of readings can often be closer to the “true” value than any individual reading is. However. just record one reading and say that it is the measured temperature at a given moment and under particular conditions. This source of uncertainty is frequently referred to as “resolution”. These statistics are quite straightforward and give us the uncertainty associated with the repeatability (or. it probably is! Such changes may be due to variations in the room temperature itself. The only information we have is a series of readings and a calculated average.9.1 2. we are imposing an identical effect by the fact that we have recorded this result to three decimal places. This would have little meaning. we can “smooth out” the effect of shortterm variations in the thermometer indication.5 2.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. could well be different. and it can be performed easily on most scientific calculators.8. If we write down a rounded result of 123. This means that we will never obtain the “true” mean value that would be revealed if we could carry out an infinite (or very large) number of measurements. and an uncertainty of ± 0. as we know that the next reading. this is often referred to as simply the standard deviation of the mean.3 It can therefore be seen that there will always be an uncertainty of ± half of the change represented by one increment of the last displayed digit. such as calibration uncertainty and secular stability. First we calculate the estimated standard deviation using the values we have measured. On some calculators it is identified n1 as s(x) or simply s. more correctly.9.8. 2.0005 will arise.4 2.2 We could. Middlesex. 2147 High Street. value. So what can be done about this? 2.9. This uncertainty cannot be evaluated using methods like those we have already considered.9. We therefore have to use a statistical approach to determine how far our calculated mean could be away from the “true” mean. we can only take a finite number of measurements. In this way. The average. We have considered what happens with finite resolution by logical reasoning.4 2. This does not only apply to digital displays. a few seconds later. 2.05 °C. because there is no background information available upon which to base our evaluation.ukas. variations in the performance of the thermometer and variations in other influence quantities. This facility is indicated on most calculators by the function key xσ .
0 °C.2 2.9.11.11. 23.ukas. We could then compare this result with that obtained when the operator is holding it in the usual manner. In this case.9.1304 °C. If we are using a liquid in glass thermometer. TW13 4UN Website: www.10 2. These effects may or may not be significant.9. In this case.10 2.9. Middlesex. This will reveal how sensitive the thermometer is to the quantity we are concerned about.3 United Kingdom Accreditation Service.2915 °C by the square root of 5. that the measurement may not be independent of the operator and that special consideration may have to be given to operator effects. This is a positive benefit of uncertainty evaluation. The relative humidity in the room can vary considerably.12 2. evaluation of uncertainty may reveal ways in which the method can be improved.8 The standard deviation of the mean is then obtained by dividing the value obtained in 2. So we have to consider whether any particular aspect of the environment could have an effect on the measurement result.11 2. This also raises a general point that is applicable to all measurements.4 °C. For example.11. UKAS Copyright exists in all UKAS publications PAGE 7 OF 82 .9 2.10.9. 5 2 . 0 . it is difficult to see how the relative humidity could significantly affect the expansion of the liquid.11 2. First. Am I warming it up? Quite possibly. so we divide 0. 2147 High Street.9 °C. or we could include a contribution to measurement uncertainty based on the results of the experiment.6 °C and 22. and significantly.7 by the square root of the number of measurements that contributed to the mean value. The sensor itself could also be affected by relative humidity.1 2.10. 23. without the operator nearby. These are 23. it is unavoidable.2915 °C. There may be heat conduction from the hand to the temperature sensor. thus giving more reliable results. Feltham. we could either improve the method so that operator effects are eliminated.236 Further information on the statistical analysis processes used for evaluation of nonrepeatability can be found in Section 4. This would yield empirical data on the effects of heat conduction and radiation.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Using the calculator function xσ . or in a variety of manners.10. If this turns out to be significant. but we will not know until an evaluation is performed. 2. Additionally. Will this affect my results? Maybe.9.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 2. we could expose the thermometer to an environment in which the temperature can be maintained constant but the relative humidity can be varied. We may have to train the operator to use the equipment in a particular way. we need means of evaluating any such effects. special experiments may be required in order to determine the significance of the effect. Special experiments may be necessary to evaluate particular effects. I am holding the thermometer in my hand. Let us try an example. There may be radiated heat from the body impinging on the sensor. Every measurement we make has to be carried out in an environment of some kind. As with other influences. This reveals a number of important issues. 23. n1 Five measurements contributed to the mean value. Suppose we record five consecutive readings with our thermometer. However. How could we do this? Some fairly basic and obvious methods come to mind.2915 0. we obtain an estimated standard deviation of 0. if we are using a digital thermometer it is quite possible that relative humidity could affect the electronics that amplify and process the signal from the sensor.2915 = 0.1 °C. we could set up the thermometer in a temperaturestable environment and read it remotely.1 2.2 2.3 2.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT
M3003  EDITION 2  JANUARY 2007
2.11.4
The significance of a particular aspect of the environment has to be considered in the light of the specific measurement being made. It is difficult to see how, for example, gravity could significantly influence the reading on a digital thermometer. However, it certainly will affect the results obtained on a precision weighing machine that might be right next to the thermometer! The following environmental effects are amongst the most commonly encountered when considering measurement uncertainty: Temperature Relative humidity Barometric pressure Electric or magnetic fields Gravity Electrical supplies to measuring equipment Air movement Vibration Light and optical reflections Furthermore, some of these influences may have little effect as long as they remain constant, but could affect measurement results when they start changing. Rate of change of temperature can be particularly important.
2.11.5
2.11.6
It can be seen by now that understanding of a measurement system is important in order to identify and quantify the various uncertainties that can arise in a measurement situation. Conversely, analysis of uncertainty can often yield a deeper understanding of the system and reveal ways in which the measurement process can be improved. This leads on to the next question… Does it matter where in the room I make the measurement? It depends what we are trying to measure! Are we interested in the temperature at a specific location? Or the average of the temperatures encountered at any location within the room? Or the average temperature at bench height? There may be further, related questions. For example, do we require the temperature at a particular time of day, or the average over a specific period of time? Such questions have to be asked, and answered, in order that we can devise an appropriate measurement method that gives us the information we require. Until we know the details of the method, we are not in a position to evaluate the uncertainties that will arise from that method. This leads to what is perhaps the most important question of all, one that should be asked before we even start with our evaluation of uncertainty: “What exactly is it that I am trying to measure?” Until this question is answered, we are not in a position to carry out a proper evaluation of the uncertainty. The particular quantity subject to measurement is known as the measurand. In order to evaluate the uncertainty in a measurement system, we must define the measurand otherwise we are not in a position to know how a particular influence quantity affects the value we obtain for it. The implication of this is that there has to be a defined relationship between the influence quantities and the measurand. This relationship is known as the mathematical model. This is an equation that describes how each influence quantity affects the value assigned to the measurand. In effect, it is a description of the measurement process. Further details about the derivation of the mathematical model can be found in Appendix D. A proper analysis of this process also gives the answer to another important question:
2.12 2.12.1
2.12.2 2.12.3
2.12.4
2.13 2.13.1
2.13.2
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 8 OF 82
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT
M3003  EDITION 2  JANUARY 2007
2.14 2.14.1
“Am I actually measuring the quantity that I thought I was measuring?” Some measurement systems are such that the result would be only an approximation to the “true” value, even if no other uncertainties were present, because of assumptions and approximations inherent in the method. The model should include any such assumptions and therefore uncertainties that arise from them will be accounted for in the analysis. Summary This section of M3003 has given an overview of uncertainty and some insights into how uncertainties might arise. It has shown that we have to know our measurement system and the way in which the various influences can affect the result. It has also shown that analysis of uncertainty can have positive benefits in that it can reveal where enhancements can be made to measurement methods, hence improving the reliability of measurement results. The following sections of M3003 explore the issues identified in this overview in more detail.
2.15 2.15.1
2.15.2
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 9 OF 82
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT
M3003  EDITION 2  JANUARY 2007
3
IN MORE DETAIL…
3.1
The Overview section of M3003 has provided an introduction to the subject of uncertainty evaluation and has explored a number of the issues involved. This section provides a more formal description of these processes, using terminology consistent with that in the GUM. A quantity (Q) is a property of a phenomenon, body or substance to which a magnitude can be assigned. The purpose of a measurement is to assign a magnitude to the measurand; the quantity intended to be measured. The assigned magnitude is considered to be the best estimate of the value of the measurand. The uncertainty evaluation process will encompass a number of influence quantities that affect the result obtained for the measurand. These influence, or input, quantities are referred to as X and the output quantity, i.e., the measurand, is referred to as Y. As there will usually be several influence quantities, they are differentiated from each other by the subscript i, So there will be several input quantities called Xi , where i represents integer values from 1 to N, N being the number of such quantities. In other words, there will be input quantities of X 1, X 2, … X N. Each of these input quantities will have a corresponding value. For example, one quantity might be the temperature of the environment – this will have a value, say 23 °C. A lowercase “x” represents the values of the quantities. Hence the value of X 1 will be x 1, that of X 2 will be x2 , and so on. The purpose of the measurement is to determine the value of the measurand, Y. As with the input uncertainties, the value of the measurand is represented by the lowercase letter, i.e. y. The uncertainty associated with y will comprise a combination of the input, or xi , uncertainties. One of the first steps is to establish the mathematical relationship between the values of the input quantities, xi , and that of the measurand, y. This process is examined in Appendix D. The values xi of the input quantities Xi will all have an associated uncertainty. This is referred to as u(xi), i.e. “the uncertainty of xi”. These values of u(xi) are known as standard uncertainties – but more on this shortly. Some uncertainties, particularly those associated with the determination of repeatability, have to be evaluated by statistical methods. Others have been evaluated by examining other information, such as data in calibration certificates, evaluation of longterm drift, consideration of the effects of environment, etc. The GUM differentiates between statistical evaluations and those using other methods. It categorises them into two types – Type A and Type B. A Type A evaluation of uncertainty is carried out using statistical analysis of a series of observations. Further details about Type A evaluations can be found in Section 4. A Type B evaluation of uncertainty is carried out using methods other than statistical analysis of a series of observations. Further details about Type B evaluations can be found in Section 5. In paragraph 3.3.4 of the GUM it is stated that the purpose of the Type A and Type B classification is to indicate the two different ways of evaluating uncertainty components, and is for convenience in discussion only. Whether components of uncertainty are classified as `random' or `systematic' in relation to a specific measurement process, or described as Type A or Type B depending on the method of evaluation, all components regardless of classification are modelled by probability distributions quantified by variances or standard deviations.
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9 3.10
3.11 3.12
PAGE 10 OF 82
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT
M3003  EDITION 2  JANUARY 2007
3.13
Therefore any convention as to how they are classified does not affect the estimation of the total uncertainty. But it should always be remembered that, in this publication, when the terms `random' and `systematic' are used they refer to the effects of uncertainty on a specific measurement process. It is the usual case that random components require Type A evaluations and systematic components require Type B evaluations, but there are exceptions. For example, a random effect can produce a fluctuation in an instrument's indication, which is both noiselike in character and significant in terms of uncertainty. It may then only be possible to estimate limits to the range of indicated values. This is not a common situation but when it occurs a Type B evaluation of the uncertainty component will be required. This is done by assigning limit values and an associated probability distribution, as in the case of other Type B uncertainties. The input uncertainties, associated with the values xi of the influence quantities Xi , arise in a number of forms. Some may be characterised as limit values within which little is known about the most likely place within the limits where the “true” value may lie. A good example of this is the numeric rounding caused by finite resolution described in paragraph 2.8. In this example, it is equally likely that the underlying value is anywhere within the defined limits of ± half of the change represented by one increment of the last displayed digit. This concept is illustrated in Figure 1. a a
3.14
3.15
3.16
probability p
xi  a
xi Figure 1
xi + a
The expectation value xi lies in the centre of a distribution of possible values with a halfwidth, or semirange, of a.
3.17 3.18
In the resolution example, a = 0.5 of a least significant digit. It can be seen from this that there is equal probability of the value of xi being anywhere within the range xi  a to xi + a, and zero probability of it being outside these limits. Thus, a contribution of uncertainty from the influence quantity can be characterised as a probability distribution, i.e. a range of possible values with information about the most likely value of the input quantity xi. In this example, it is not possible to say that any particular position of xi within the range is more or less likely than any other. This is because there is no information available upon which to make such a judgement. The probability distributions associated with the input uncertainties are therefore a reflection of the available knowledge about that particular quantity. In many cases, there will be insufficient information available to make a reasoned judgement and therefore a uniform, or rectangular, probability distribution has to be assumed. Figure 1 is an example of such a distribution. If more information is available, it may be possible to assign a different probability distribution to the value of a particular input quantity. For example, a measurement may be taken as the difference in readings on a digital scale – typically, the zero reading will be subtracted from a reading taken further up the scale. If the scale is linear, both of these readings will have an associated rectangular distribution of identical size. If two identical rectangular distributions, each of magnitude ± a, are combined then the resulting distribution will be triangular with a semirange of ± 2a.
3.19
3.20
3.21
United Kingdom Accreditation Service, 2147 High Street, Feltham, Middlesex, TW13 4UN Website: www.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500
© United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications
PAGE 11 OF 82
Harris and [3] Warner have shown that a symmetrical Ushaped probability distribution arises from this effect. will usually yield a Gaussian or normal distribution.a xi Figure 3 xi + a Ushaped distribution. TW13 4UN Website: www. This means that a cosine function characterises the probability distribution for the uncertainty. yields a triangular distribution with a semirange of ± 2a. In this example. associated with RF mismatch uncertainty. Middlesex. the distribution has been evaluated from a theoretical analysis of the principles involved.ukas. when making measurements of radiofrequency power an uncertainty arises due to imperfect matching between the source and the termination. 3.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. each with semirange limits of ± a. For example. For this situation. performed by statistical methods. UKAS Copyright exists in all UKAS publications PAGE 12 OF 82 . Feltham. xi + 2a xi . Further details on this process can be found in Section 4. 3.23 An evaluation of the effects of nonrepeatability.2a xi Figure 2 Combination of two identical rectangular distributions. xi is likely to be close to one or other of the edges of the distribution. 2147 High Street.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 probability p xi . United Kingdom Accreditation Service.22 There are other possible distributions that may be assigned. The imperfect match usually involves an unknown phase angle.
they will combine to form a reasonable approximation to the normal distribution. In practice.24 When a number of distributions of whatever form are combined it can be shown that.ukas. probability distribution. It will be therefore be necessary to express the input uncertainties in terms that.25 When the input uncertainties are combined. Then. the standard uncertainty is given by u i i . Figure 4 The normal. The exceptional case arises when one contribution to the total uncertainty dominates.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 3. When it is possible to assess only the upper and lower bounds of an error.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. in this circumstance the resulting distribution departs little from that of the dominant contribution. Middlesex. 3. The normal distribution is described in terms of a standard deviation. are combined and the conditions of the Central Limit Theorem are met. a rectangular probability distribution should be assumed for the uncertainty associated with this error. The shaded area represents ± 1 standard deviation from the centre of the distribution.26 3. then clearly the resulting distribution will also be normal. Table 1 gives the expressions for this and for x 3 other situations. the resulting probability distribution tends to the normal form in accordance with the [4] Central Limit Theorem . As some of the input uncertainties are expressed as limit values (e. The importance of this is that it makes it possible to assign a confidence level in terms of probability to the combined uncertainty. This is obtained when a number of distributions.. This corresponds to approximately 68% of the area under the curve. if three or more distributions of similar magnitude are present. when combined. apart from in exceptional cases. UKAS Copyright exists in all UKAS publications PAGE 13 OF 82 . Feltham. of any form. a normal distribution will usually be obtained. will cause the resulting normal distribution to be expressed at the one standard deviation level. NOTE: If the dominant contribution is itself normal in form. TW13 4UN Website: www. or Gaussian. like the example in Figure 4. the rectangular distribution).g.27 a limit. which is known as a standard uncertainty and is referred to as u(xi). The size of the distribution is described in terms of a standard deviation. United Kingdom Accreditation Service. 2147 High Street. some processing is needed to convert them into this form. if a i is the semirange 3.
1°C.029°C. Thus the corresponding standard uncertainty ai 1 . with no probability of being outside the 3°C limits.3%. TW13 4UN Website: www. The numeric rounding caused by finite resolution will have semirange limits of 0. In such cases. will have been used to obtain this expanded uncertainty from the combination of standard uncertainties. therefore no further processing is required.3 will be u i x 0. It is therefore necessary to divide the expanded uncertainty by the same coverage factor to obtain the standard uncertainty. k = 2 and for a coverage probability of 99%. therefore the standard uncertainty for ai 3 its temperature is u i x 1.g.58.732 A mismatch uncertainty associated with the calibration of an RF power sensor has been evaluated as having semirange limits of 1. k = 2. k. Thus the corresponding ai 0. 2147 High Street.92% 1 .449 A statistical evaluation of repeatability gives the result in terms of one standard deviation. Some manufacturers’ specifications are quoted at a given coverage probability (sometimes referred to as confidence level). It is reasonable to assume a triangular distribution. 95% or 99%. A coverage factor. Ushaped a u i i x 2 Triangular a u i i x 6 Normal (from repeatability evaluation) u i s x q Normal (from a calibration certificate) ux i U k Normal (from a manufacturer’s specification) u i x Tolerance limit k United Kingdom Accreditation Service. If a coverage probability is not stated then a rectangular distribution should be assumed. For a coverage probability of 95%. UKAS Copyright exists in all UKAS publications PAGE 14 OF 82 . A calibration certificate normally quotes an expanded uncertainty U at a specified. high coverage probability.ukas. a normal distribution can be assumed and the tolerance limit is divided by the coverage factor k for the stated coverage probability. e.05 standard uncertainty will be u i x 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 Table 1 Assumed probability distribution Rectangular Expression used to obtain the standard uncertainty Comments or examples a u i i x 3 A digital thermometer has a least significant digit of 0. 3 1.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.2°C 6 2 .414 2 A tensile testing machine is used in a testing laboratory where the air temperature can vary randomly but does not depart from the nominal value by more than 3°C. The machine has a large thermal mass and is therefore most likely to be at the mean air temperature. Feltham. Middlesex.05°C.
In this example. In other words. So.0 m and the measured angle is 37°. Because the gauge blocks have a significant temperature coefficient of expansion.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 3. It is given by 3. it describes how the output estimate y varies with a corresponding small change in an input estimate xi.33 Example The height h of a flagpole is determined by measuring the angle obtained when observing the top of the pole at a specified distance d. 2147 High Street. If the measured distance is 7.275 m. It is important to choose the magnitude of the f x change Δ i around xi carefully. one to one. The sensitivity coefficient simply describes how sensitive the result is to a particular influence quantity. For example.29 3.5 x 10 per °C. It should be balanced between being sufficiently large to obtain adequate x numerical accuracy in Δ and sufficiently small to provide a mathematically sound approximation to the f partial derivative. 3. while keeping all other inputs constant. a sensitivity coefficient is required. Φ d United Kingdom Accreditation Service. evaluated at the input estimates xi. this figure can be used as the sensitivity coefficient. and why it is necessary to know the functional relationship between the influence quantities and the measurand. In other words. Thus h = d tan Φ.ukas.31 In other words. It is the partial derivative of the model function f with respect to X i. TW13 4UN Website: www. In order to translate the temperature uncertainty into an uncertainty in length units. and noting the change in the output estimate. the estimated height is 7. 3. Feltham. there is an uncertainty that arises in their length due to an uncertainty in temperature units.0 x tan (37) = 5.32 The calculations required to obtain sensitivity coefficients by partial differentiation can be a lengthy process. Middlesex. This approach can also be used if f is known but the determination of the partial derivatives is likely to be difficult. it is necessary to know how sensitive the length of the gauge block is to temperature. h = f(d) = d tan Φ. they may be entirely different units altogether. relationship with it. The following example illustrates this. in this case. If the functional relationship is not known for a particular measurement system the sensitivity coefficients can sometimes be obtained by the practical approach of changing one of the input variables by a known amount. UKAS Copyright exists in all UKAS publications PAGE 15 OF 82 . The sensitivity coefficient associated with each input estimate x i is referred to as ci . a dimensional laboratory may use steel gauge blocks for calibration of measuring tools. A more straightforward approach is to replace the partial derivative ∂/ ∂i by the quotient Δ / Δ i .28 The quantities Xi that affect the measurand Y may not have a direct.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. where f x f x Δ is the change in f resulting from a change Δ i in xi. the steel used in the manufacture of gauge blocks has a temperature coefficient of 6 expansion of approximately +11. particularly when there are many input contributions and uncertainty estimates are needed for a range of values. A significant influence quantity is temperature. Both h and d are in units of length but are related by tan Φ.30 3. Indeed.
192 metre/degree 1 Standard uncertainty ui (y) 3. As the input uncertainties had been expressed in terms of a standard uncertainty.39 In other words.ukas.075 coefficient is therefore 0.40 0.5).0554 2 0. these squared values are added and the square root is taken. The sensitivity 0. the uncertainties have to be combined in order to give a single value of uncertainty to be associated with the estimate y of the measurand Y.0863 m Combined standard uncertainty uc(y) = NOTE 0 .75 = 0. are squared.179 m and 5.05 = 0. it is assumed that the repeatability of the process has been evaluated by making repeat measurements of the flagpole height.75 .5° 0. A change of ± 0. The combined standard uncertainty is calculated as follows: 3.075 m in the output estimate y. this combined standard uncertainty takes the form of a normal distribution.5° in the input quantity xi has resulted in a change of ± 0.1 m 0.05 2 The values in the righthand column of the table are now scaled in accordance with the effect of the corresponding input quantity on the measurand and are expressed as a standard uncertainty.0 tan (36.75 0.0433 m 0 . expressed in terms of the measurand. The sensitivity coefficient is therefore 0 .1) tan (37). For the purposes of the example.1) tan (37) and (7. between 5.192 metre per degree.0554 m 0 0.1 m then the estimate of h could be anywhere between (7.5°. United Kingdom Accreditation Service. i.350 m. 2147 High Street.41 In accordance with the Central Limit Theorem. using the data from the measurement of the flagpole height described on page 15. Source of uncertainty Distance from flagpole Angle measurement Repeatability Value Probability distribution Rectangular Rectangular Normal Divisor Sensitivity coefficient 0.05 m √ 3 √ 3 1 0 . ± 0. 0 .371 m.5 3 . Middlesex.e.5) and 7.0433 2 0 . i. then the estimate of h could be anywhere between 7. giving an estimated standard deviation of the mean of 0.1 3 0. the individual standard uncertainties.0 + 0. Feltham.096 m 0.0 tan (37. 0 . This is known as the combined standard uncertainty and is given the symbol uc(y). See Section 4 for further details about the evaluation of repeatability. say.05 metres. A change of ± 0.34 If the uncertainty in d is.096 in the output estimate y.e.1 m in the input quantity x i has resulted in a change of ± 0. If the uncertainty in Φ is ± 0. between 5.200 m and 5.37 3. UKAS Copyright exists in all UKAS publications PAGE 16 OF 82 . 3.36 3.192 = 0.05 m 1 1 = 0.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.35 3.1 Similar reasoning can be applied to the uncertainty in the angle Φ. as illustrated in Figure 5.0 – 0. the resulting normal distribution is expressed as one standard deviation.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 3. An example of this process is presented below.5 Once the standard uncertainties xi and the sensitivity coefficients ci have been evaluated.38 u c y 2 2 2 x u y c i u i i N N i 1 i 1 (1) 3. TW13 4UN Website: www.
For example.0863 = 0. This means that there is about 68% confidence that the measured value y lies within the stated limits.43 In accordance with generally accepted international practice. There may also be situations where a normal distribution can be assumed.44 (2) y + 0. 2147 High Street. However.64 1.00 2. Middlesex. The GUM recognises the need for providing a high level of confidence – referred to herein as coverage probability . 3. 1 standard deviation encompasses 68.58 3. TW13 4UN Website: www. in safetycritical situations a higher coverage probability may be more appropriate. the difference is not generally significant since.47 United Kingdom Accreditation Service.45 Example: The measurement of the height of the flagpole had a combined standard uncertainty uc(y) of 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 y 5. UKAS Copyright exists in all UKAS publications PAGE 17 OF 82 . 3. Such situations are described in Appendix B and Appendix C.42 For a normal distribution.00 3. NOTE: A coverage factor of k = 2 actually provides a coverage probability of 95.45% 99% 99.0863 m.46 3.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. There may be situations where a normal distribution cannot be assumed and a different coverage factor may be needed in order to obtain a coverage probability of approximately 95%. thus the expanded uncertainty is given by U = k u c(y). This value of k will give a coverage probability of approximately 95%. The table below gives the coverage factor necessary to obtain various levels of confidence for a normal distribution. Coverage probability p 90% 95% 95. U. 3. but a different coverage probability is required.ukas. The figures shown relate to the example discussed above. it is recommended that a coverage factor of k = 2 is used to calculate the expanded uncertainty. For convenience this is approximated to 95% which would relate to a coverage factor of k = 1. which is obtained by multiplying the combined standard uncertainty by a coverage factor. in practice.173 m.0863 m 3. The coverage factor is given the symbol k.27% of the area under the curve.275 m y – 0. assuming a normal distribution. Feltham.0863 m Figure 5 The measured value y is at the centre of a normal distribution with a standard deviation equal to u c(y). Hence the expanded uncertainty U = k uc (y) = 2 x 0.45% for a normal distribution.73% Coverage factor k 1.96. the coverage probability is usually based on conservative assumptions and approximations to the true probability distributions.96 2.associated with an uncertainty and uses the term expanded uncertainty.
2147 High Street. it is necessary to make an estimate of the value of σthat would be obtained were this possible. s(qj ).5 4.1 4. However. an estimate. the arithmetic mean or average of the results should be calculated. another set of values. Similarly. Equation (4) gives the standard deviation for the samples actually selected. If the process is repeated. and sometimes on the person making the measurement. which is given by σ 1 n q n j 1 j q 2 (4) 4. UKAS Copyright exists in all UKAS publications PAGE 18 OF 82 .THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 4 4. the random component of uncertainty may not be significant in relation to other contributions to uncertainty. therefore an estimate has to be made of the possible error from the “true” mean. these are not the only values that could have been sampled. It is obtained from the estimated standard deviation of the population by the expression: s j q s q (6) n United Kingdom Accreditation Service. rather than of the whole population of possible samples. For some measurements. can be made of the standard deviation σof the whole population of possible values of the measurand from the relation 4. Middlesex. However. This uncertainty is referred to as the experimental standard deviation of the mean. these mean values approach the central limit of a distribution of all possible values.4 This expression yields the standard deviation σof the particular set of values sampled. the method.2 TYPE A EVALUATION OF STANDARD UNCERTAINTY If an uncertainty is evaluated by statistical analysis of a series of observations. The mean value itself therefore has uncertainty. with different values of q and σ will be obtained.3 q j 1 n j q 2 3 qn q q 1 n (3) The spread in the results gives an indication of the repeatability of the measurement process. which depends on various factors. the mean value obtained is less likely to be the same as that which would be obtained if a very large number of measurements could be taken.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. including the apparatus used. it is known as a Type A evaluation. A good description of this spread of values is the standard deviation σof the n values that comprise the sample. When there is a significant spread in a sample of measurement results. As it is impractical to capture all values that are available. This probability distribution can often be assumed to have the normal form. from the results of a single sample of measurements. Feltham. It is nevertheless desirable for any measurement process that the relative importance of random effects be established.8 The mean value q will have been derived from a finite number n of samples and therefore its value will not be the exact mean that would have been obtained if an infinite number of samples could have been taken. A Type A evaluation will normally be used to obtain a value for the repeatability or randomness of a measurement process. . If there are n independent repeated values for a quantity Q then the mean value q is given by 1 q n 4. For large values of n. TW13 4UN Website: www.ukas.7 s qj 1 n 1 q n j 1 j q 2 (5) 4.6 4.
i. It is also recommended that data obtained from prior assessment should be regularly reviewed. the estimated standard deviation s(q j) is given by 4.9 Example: Four measurements were made to determine the repeatability of a measurement system.88 2. 2 (7) where m is the number of readings considered in the previous evaluation. (8) A previous estimate of standard deviation can only be used if there has been no subsequent change in the measurement system or procedure that could have an effect on the repeatability.42 3.386 q The experimental standard deviation of the mean s q 0. u i s x q .11 s qj 4. In these cases a more reliable estimate of the standard deviation of a measurement system may be obtained from data obtained previously.42. and to rely on a previous assessment of the repeatability of similar devices. The results obtained were 3. 4. United Kingdom Accreditation Service. this is the reason that a large number of readings in a prior evaluation can give a more reliable estimate when only a few measurements can be made during the routine procedure. and n in Equation (6) is then 1. TW13 4UN Website: www. however.ukas. In such cases. 2. Whenever possible at least two measurements should be made as part of the procedure.13 4. Feltham.14 The standard uncertainty is then the standard deviation of the mean. which is not typical of the measurement system. If an apparently excessive spread in measurement values is found. based on a larger number of readings.e.99 and 3. 3.193 n 4 For information on the use of calculators to calculate s(qj).386 n j 2 s j 0. 2147 High Street. 4. Indeed.12 1 m q q m j j 1 1 .99 3.365 4 The estimated standard deviation s q j 1 n j 1 1 q q = 0. it is acceptable for a single measurement to be made even though it is known that the system has imperfect repeatability.88.10 It may not always be practical or possible to repeat the measurement many times during a test or a calibration. UKAS Copyright exists in all UKAS publications PAGE 19 OF 82 . Degrees of freedom are discussed further in Appendix B. This procedure must be treated with caution because the reliability of a previous assessment will depend on the number of devices sampled and how well this sample represents all devices.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 4. where m is the number of measurements in the prior evaluation. Of course. The standard deviation s qj of the mean s q . NOTE The degrees of freedom under such circumstances are m – 1. 1 The mean value q n q j 1 n j 3 . the cause should be investigated and resolved before proceeding further.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. n being the number of measurements contributing to the reported mean n value. see Paragraph P4.17. Middlesex. when only one measurement is made on the device being calibrated a value of s(qj) must have been obtained from prior measurements.17 = 3.
and usually are. there can be. Having identified all the possible systematic components of uncertainty based as far as possible on experimental data or on theoretical grounds. The equipment or item being measured. The effects of such mistakes cannot readily be included in the evaluation of uncertainty.5 United Kingdom Accreditation Service. the convention is that an error is given a positive sign if the measured value is greater than the conventional true value. corrections should be made for errors revealed by calibration or other sources. for a reference instrument.e. In evaluating the components of uncertainty it is necessary to consider and include at least the following possible sources: (a) The reported calibration uncertainty assigned to reference standards and any drift or instability in their values or readings. UKAS Copyright exists in all UKAS publications PAGE 20 OF 82 . and any drift or instability in their values or readings. for example its resolution and any instability during the measurement. 2147 High Street. to simplify the measurement process it may be convenient to treat such an error. The most important of these systematic components. other important contributions to systematic errors in measurement that arise in the equipment user's own laboratory.4 5.1 TYPE B EVALUATION OF STANDARD UNCERTAINTY It is probable that systematic components of uncertainty. The effects of environmental conditions on any or all of the above. transcription errors. they should be characterised in terms of standard uncertainties based on the assessed probability distributions.22.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. i. The successful identification and evaluation of these contributions depends on a detailed knowledge of the measurement process and the experience of the person making the measurements. These distributions and sample calculations are presented in detail in paragraphs 3. those that account for errors that remain constant while the measurement is made. Whenever possible. Feltham. It should be noted that the anticipated longterm performance of the item being calibrated is not normally included in the uncertainty evaluation for that calibration.ukas. and faults in software designed to control or report on a measurement process. Variability between different staff carrying out the same type of measurement. The correction for error involves subtracting the error from the measured value. The probability distribution of an uncertainty obtained from a Type B evaluation can take a variety of forms but it is generally acceptable to assign welldefined geometric shapes for which the standard uncertainty can be obtained from a simple calculation.2 5. when it is small compared with other uncertainties. The operational procedure. Middlesex.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 5 5. will often be the imported uncertainties associated with its own calibration. 5. TW13 4UN Website: www. including ancillaries such as connecting leads etc.. The calibration of measuring equipment. Common examples are errors in the corrections applied to values. The need for the utmost vigilance in preventing mistakes cannot be overemphasised. as if it were a systematic uncertainty equal to (±) the uncorrected error magnitude.3 (b) (c) (d) (e) (f) 5. On occasions.15 to 3. will be estimated from Type B evaluations. However.
Middlesex.2 6. For the purposes of this document "approximately" is interpreted as meaning sufficiently close that any difference may be considered insignificant.5 6. It is therefore recommended that the expanded uncertainty be rounded to two significant figures. In cases where the procedure of Appendix B has been followed the actual value of the coverage factor should be substituted for k = 2 and the following statement used: "The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = XX. However if there is a significant difference between the upper and lower values then they should be evaluated and reported separately. which for a tdistribution with veff = YY effective degrees of freedom corresponds to a coverage probability of approximately 95%.3 6. The uncertainty evaluation has been carried out in accordance with UKAS requirements".ukas.7 6. If uncertainty is being reported as an analytical expression. In the special circumstances where a dominant nonGaussian Type B contribution occurs refer to Appendix C. However there may be situations where the upper and lower uncertainty values are different. The numerical value of the measurement result should in the final statement normally be rounded to the least significant figure in the value of the expanded uncertainty assigned to the measurement result. 6.4 6. parts per million (ppm). TW13 4UN Website: www. The number of figures in a reported uncertainty should always reflect practical measurement capability.9 United Kingdom Accreditation Service. for example if cosine errors are involved.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 6 6. 2147 High Street.1 REPORTING OF RESULTS After the expanded uncertainty has been calculated for a coverage probability of 95% the value of the measurand and expanded uncertainty should be reported as y ± U and accompanied by the following statement of confidence: "The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2. refer to Appendix L . In view of the process for estimating uncertainties it is seldom justified to report more than two significant figures. Uncertainties are usually expressed in bilateral terms (±) either in units of the measurand or as X relative values.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications PAGE 21 OF 82 . providing a coverage probability of approximately 95%.8 6. etc. using the normal rules of rounding.6 6. Rounding should always be carried out at the end of the process in order to avoid the effects of cumulative rounding errors. for example as a percentage (%). Feltham. 1 in 10 . The uncertainty evaluation has been carried out in accordance with UKAS requirements". If such differences are small then the most practical approach is to report the expanded uncertainty as ± the larger of the two.
It will be assumed that the unknown weight. 2147 High Street. however. General case Example K4: Calibration of a weight of nominal value 10 kg of OIML Class M1 7. Middlesex. Feltham. can be obtained from the following relationship: WX = WS + DS + δd + δ + Ab I C 7. … xN ) See Appendix D for details. It is not normal practice to apply corrections for this class of weight and the comparator has no measurable linearity error. TW13 4UN Website: www.1 If possible determine the mathematical relationship between values of the input quantities and that of the measurand: y = f (x1. the process for testing activities is unchanged.2 Identify all corrections that have to be applied to the results of measurements of a quantity (measurand) for the stated conditions of measurement. therefore: Drift of standard mass since last calibration: Correction for air buoyancy: Linearity correction: Effect of least significant digit resolution: 0 0 0 0 United Kingdom Accreditation Service. W X. uncertainties for these contributions have been determined. The left hand column gives the general case while the right hand column indicates how this relates to example K4 in Appendix K. Although this example relates to a calibration activity. UKAS Copyright exists in all UKAS publications PAGE 22 OF 82 .THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 7 STEP BY STEP PROCEDURE FOR EVALUATION OF MEASUREMENT UNCERTAINTY The following is a guide to the use of this code of practice for the treatment of uncertainties.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. x 2.
2147 High Street. UKAS Copyright exists in all UKAS publications PAGE 23 OF 82 . Calculate the standard uncertainty for each component of uncertainty. Middlesex. obtained from Type B evaluations.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 General case Example K4: Calibration of a weight of nominal value 10 kg of OIML Class M1 7.020 WS = 0. 30 u 1 u S x W = 15 mg 2 30 u 2 u S x D = 17. Seek prior experimental work or theory as a basis for assigning uncertainties and probability distributions to the systematic components of uncertainty. For assumed rectangular distributions: Source of uncertainty Limit (mg) Distribution WS Calibration of std.4 Use prior knowledge or make trial measurements and calculations to determine if there is going to be a random component of uncertainty that is significant compared with the effect of the listed systematic components of uncertainty. mass DS Drift of standard mass δ C Comparator linearity δ Ab Air buoyancy δ Id Resolution effects 30 Normal (k = 2) 30 Rectangular 3 Rectangular 10 Rectangular a u i i x 3 For assumed triangular distributions: a ux i i 6 10 Triangular Then: For assumed normal distributions: U ux i k or consult other documents if the assumed probability distribution is not covered in this publication. 7.3 List systematic components of uncertainty associated with corrections and uncorrected systematic errors treated as uncertainties.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.025 0. from which the mean difference was calculated: 0 .08 mg 7. Random components of uncertainty also have to be considered as input quantities.ukas. Feltham.5 If a random component of uncertainty is significant make repeated measurements to obtain the mean from Equation (3): n 1 q qj n j 1 Three measurements were made of the difference between the unknown weight and the standard weight.020 g 3 United Kingdom Accreditation Service. as in Table 1.77 mg 3 u 5 u Id x δ 10 6 = 4.015 0. From previous knowledge of the measurement system it is known that there is a significant random component of uncertainty. TW13 4UN Website: www.32 mg 3 u 3 u C x δ 3 3 = 1.73 mg 10 u 4 u x Ab = 5.
com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. The standard deviation was determined from 10 measurements using the conventional bracketing technique and was calculated.7 Even when a random component of uncertainty is not significant.8 Derive the standard uncertainty for the above Type A evaluation from Equation (8): This is then the standard uncertainty for the Type A evaluation: u i s x q u 6 u WR s WR = 5.ukas. Middlesex. 2147 High Street.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 General case Example K4: Calibration of a weight of nominal value 10 kg of OIML Class M1 7. Since the number of determinations taken when calibrating the unknown weight was 3 this is the value of n that is used to calculate the standard deviation of the mean using Equation (6): s 8 . using Equation (7): s j q 1 m 1 q q m j j 1 2 A previous Type A evaluation had been made to determine the repeatability of the comparison using the same type of 10 kg weights. using Equation (5). 7.0 mg n 3 s j q s q n where m is the number of readings used in the prior evaluation and n is the number of readings that contribute to the mean value. where possible check the instrument indication at least once to minimise the possibility of unexpected errors.7 W sWX = 5. TW13 4UN Website: www. 7.0 mg x United Kingdom Accreditation Service. Feltham. to be 8.6 Either calculate the standard deviation of the mean value from Equations (5) and (6): s qj 1 n qj q n j 1 1 2 s j q s q n or refer to the results of previous repeatability evaluations for an estimate of s(q j) based on a larger number of readings.7 mg. UKAS Copyright exists in all UKAS publications PAGE 24 OF 82 .
7.9 Calculate the combined standard uncertainty for uncorrelated input quantities using Equation (1) if absolute values are used: u c y 2 2 2 x u y . if there is a significant random contribution evaluated from a small number of readings.10 If correlation is suspected use the guidance in paragraph D3 or consult other referenced documents.55 mg = 49. TW13 4UN Website: www. and the functional relationship between the input quantities and the measurand is a linear summation. i. United Kingdom Accreditation Service.e.025 g ± 0. providing a coverage probability of approximately 95%. therefore Equation (1) can be used to calculate the combined standard uncertainty: u X 15 . The uncertainty evaluation has been carried out in accordance with UKAS requirements.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.049 g. c i u i i where ci is the N N i 1 i 1 The units of all standard uncertainties are in terms of those of the measurand. The measured value of the 10 kg weight is 10 000. milligrams.11 Either calculate an expanded uncertainty from Equation (2): U = k·uc (y) or.08 . 7.73 . N 2 u c y y 7. or a known sensitivity f x coefficient.32 4.ukas.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 General case Example K4: Calibration of a weight of nominal value 10 kg of OIML Class M1 7. Alternatively use Equation (11) if the standard uncertainties are relative values: u p x i xi i . Middlesex. None of the input quantities is considered to be correlated to any significant extent.12 Report the result and the expanded uncertainty in accordance with Section 6. U = 2 x 24.55 mg. 2147 High Street.0 W 17 1 5 5 2 2 2 2 2 2 = 24. Feltham. The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2. where pi are known i 1 positive or negative exponents in the functional relationship. UKAS Copyright exists in all UKAS publications PAGE 25 OF 82 .10 mg It was not necessary to use Appendix B to determine a value for kp . partial derivative i .77 . In fact the effective degrees of freedom of u(WX ) are greater than 5000 which gives a value for k95 = 2.00. therefore all the sensitivity coefficients are unity (c i =1). use Appendix B to calculate a value for kp and use this value to calculate the expanded uncertainty.
for example. Feltham. equipment and facilities that were the subject of the assessment. Where necessary the laboratory's schedule of accreditation will include a remark that describes the conditions under which the best measurement capability can be achieved. In order to promote harmony between accredited laboratories and between accreditation bodies the EA has adopted the following definition of best measurement capability: The smallest uncertainty of measurement a laboratory can achieve within its scope of accreditation. UKAS Copyright exists in all UKAS publications PAGE 26 OF 82 . no device is perfect and so the concept of “nearly ideal” is used in association with the evaluation of a BMC. It may also be the case that an accredited laboratory can achieve a particular uncertainty if conditions are optimum but cannot achieve this uncertainty routinely. realize. Middlesex. In other words "best measurement capability" is the smallest uncertainty a laboratory can achieve when performing more or less routine calibrations on a nearly ideal device being calibrated. which usually requires the use of a coverage factor of k = 2. However. using the procedures. conserve or reproduce a unit of that quantity or one or more of its values. very low voltage reflection coefficient etc. An accredited laboratory is not permitted to report an uncertainty smaller than their accredited best measurement capability but may report an equal or larger uncertainty. The properties of these devices that are considered to be nearly ideal will depend on the field of calibration but may include an instrument with very low random fluctuations. negligible temperature coefficient. when performing more or less routine calibrations of nearly ideal measurement standards intended to define. or when performing more or less routine calibrations of nearly ideal measuring instruments designed for the measurement of that quantity. Since the magnitude of the uncertainty reported on a certificate of calibration will often depend on properties of the device being calibrated any definition of best measurement capability should not include uncertainties that are dependent on this device. The best measurement capability should be calculated according to the procedures given in this document and should normally be quoted as an expanded uncertainty at a coverage probability of 95%. 2147 High Street.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX A BEST MEASUREMENT CAPABILITY A1 Best measurement capability (BMC) is a term normally used to describe the uncertainty that appears in an accredited calibration laboratory's schedule of accreditation and is the uncertainty for which the laboratory has been accredited using the procedure that was the subject of assessment.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Where a smaller uncertainty can be achieved by. A2 A3 A5 United Kingdom Accreditation Service. By "more or less routine calibrations" it is meant that the laboratory shall be able to achieve the stated capability in the normal work that it performs under its accreditation and.ukas. taking a large number of readings this should be considered when arriving at the budget for the best measurement capability and would therefore be within the "more or less routine" conditions. by implication. A4 A nearly ideal device is one that is available but does not necessarily represent the majority of devices that the laboratory may be asked to calibrate. TW13 4UN Website: www. if this is the case. The uncertainty budget that is intended to demonstrate the best measurement uncertainty should still include contributions from the properties of the device being calibrated that are considered to be nearly ideal but the value of the uncertainty may be entered as zero or a negligible value.
This is because they will maintain their own reference standards. A7 A8 United Kingdom Accreditation Service. 2147 High Street. UKAS Copyright exists in all UKAS publications PAGE 27 OF 82 . the implication is that the laboratory is uncomfortable in some way about the magnitude of the expanded uncertainty.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 A6 It is sometimes the case that a laboratory may wish to be accredited for a measurement uncertainty that is larger than it can actually achieve. If this is smaller than the uncertainty the laboratory wishes to be accredited for and report on their certificates of calibration. Guidance about the expression of uncertainty over a range of values is presented in Appendix L. If the principles of this document are followed when constructing the uncertainty budget the resulting expanded uncertainty should be a realistic representation of the laboratory’s measurement capability. It can be the case that some calibration laboratories offer a best measurement capability with very small uncertainties but these are not routinely offered for everyday calibrations. In some cases the best measurement capability quoted in a laboratory's schedule has to cover a two (or more) dimensional range of measured values. or may be described as an expression.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. but use subsidiary equipment – often automated – for routine work. Middlesex. If this is the case then the contributions to the uncertainty budget should be reviewed and consideration given to making more conservative allowances as necessary.ukas. TW13 4UN Website: www. Feltham. upon which the BMC is based. such as different levels and frequencies. The contract review arrangements between the laboratory and its customer should define the level of service being offered. In these cases the best measurement capability may be given as a range of uncertainties appropriate to the upper and lower values of the uncertainty that has been calculated for the range of the quantity. and it may not be practical to give the actual uncertainty for all possible values of the quantities.
a greater proportion of the values lie outside the region y – k to y + k. An example of the tdistribution is superimposed. In these situations the value of k. B2 y– k y Figure 6 y+k In Figure 6. of the combined standard uncertainty uc (y). using dashed lines. of the individual standard uncertainties u i(y). which could result in the coverage probability being significantly less than 95% if a coverage factor of k = 2 is used. A specified proportion p of the values under the curve are encompassed between y – k and y + k. where p is the confidence probability. Up. U. where n is i. vi . if the procedure followed for making the measurements is well established and if the Type A evaluations are obtained from a sufficient number of observations then the use of a coverage factor of k = 2 will mean that the expanded uncertainty. and a smaller proportion lie inside this region. is obtained by evaluating the effective degrees of freedom of u c(y) and obtaining the corresponding value of tp . Feltham. However.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. 2147 High Street. ν for contributions obtained from Type A evaluations are n . will provide an interval with a coverage probability close to 95%. An increased value of k is therefore required to restore the original coverage probability. Middlesex. in some cases it may not be practical to base the Type A evaluation on a large number of readings.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX B DERIVING A COVERAGE FACTOR FOR UNRELIABLE INPUT QUANTITIES B1 In the majority of measurement situations it will be possible to evaluate Type B uncertainties with high reliability. or more correctly kp. the number of readings used to evaluate s(q j). veff . B3 In order to obtain a value for kp it is necessary to obtain an estimate of the effective degrees of freedom. This value of kp will give an expanded uncertainty. that maintains the coverage probability at approximately the required level p. TW13 4UN Website: www. The GUM recommends that the WelchSatterthwaite equation is used to calculate a value for veff based on the degrees of freedom. from the tdistribution table. and hence kp. UKAS Copyright exists in all UKAS publications PAGE 28 OF 82 . United Kingdom Accreditation Service.1. For the tdistribution. should be based on a tdistribution rather than a normal distribution. Furthermore. This new coverage factor. therefore ν eff u c y 4 N 4 i 1 u i y ν i (9) B4 The degrees of freedom.ukas. the solid line indicates the normal distribution. kp . This is because the distribution tends to normality as the number of observations increases and k = 2 corresponds to 95% confidence for a normal distribution.
11 2.78 1.09 2. Degrees of freedom ν 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 25 30 35 40 45 50 100 ∞ Values of tp(ν from the tdistribution for degrees of freedom νthat define an interval ) that encompasses specified fractions p of the corresponding distribution p = 68.03 1.78 2.31 2.85 3.23 3.18 2.12 2.90 4.28 4.88 2.ukas. their value is known with a very high degree of reliability.09 1.72 2.960 13.09 2.31 2.576 235.69 3. the values corresponding to p = 95.03 1.27 3.08 1.005 1.52 2.645 12.04 1.80 1.09 2.11 3.07 2.03 2.01 1. for various levels of confidence p. gives values for kp.25 2.000 63.20 1.06 2.04 1.92 5.87 2.60 4.23 2.26 2.01 1.65 2. This is illustrated in the example in paragraph B10.75 1.85 2.96 3.06 2.01 1.90 2.53 4.04 1.20 2.11 2.32 1.86 1.89 1.54 3.660 1.68 2.000 B6 B7 United Kingdom Accreditation Service.984 1.32 2.71 3.10 2.20 3. Unless otherwise specified.66 9. a Type B contribution may have come from a calibration certificate as an expanded uncertainty based on a tdistribution rather than a normal distribution.01 2.13 2.98 2.75 1. However.025 2.48 3.05 2. Having obtained a value for veff .01 1.16 2. then the process using the WelchSatterthwaite formula simplifies.45% p = 99% p = 99.05 1.14 1.03 1.13 2.64 3.18 2.97 4.45 2.03 1.077 3.72 1.03 3.71 4. as described in this Appendix.626 2. the tdistribution table is used to find a value of kp .92 2. as all the terms relating to the type B uncertainties become zero.92 2.36 3.43 2.50 3.73 1.33 3.14 2. reproduced below.70 1. Feltham. ν.09 3.83 1.68 1.02 1.68 1. If this is the case.16 3.13 2. This is an example of a Type B uncertainty that does not have infinite degrees of freedom.51 4.05 1.16 2.80 19.31 2.04 2.23 2.86 2.73% 1.05 3.28 2. of Type B uncertainty contributions as infinite.79 2.73 1.22 6. For this eventuality the degrees of freedom will be as quoted on the calibration certificate.06 2.15 2.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.35 2.000 6.20 2.62 5.25 3.75 2. UKAS Copyright exists in all UKAS publications PAGE 29 OF 82 .69 2.36 2. i that is.17 2.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 B5 It is often possible to take the degrees of freedom.59 3.18 2.01 1.02 2.81 1.21 2.77 1.01 1. or it can be obtained from the tdistribution table for the appropriate value of k95 .74 1.14 2.70 2.03 1.01 2.03 1. This table.57 2. TW13 4UN Website: www.11 1.95 2.94 1. Middlesex.53 3.68 1.30 3.07 1. and there is only one contribution obtained from a Type A evaluation. 2147 High Street.51 3.45 3.76 3.76 1.06 1.37 2.21 9.45% should be used.84 4.42 3.02 1.17 3.70 1.27% p = 90% p = 95% p = 95.18 3.84 1.71 1.
had a value of 5.7 x 2. the next lower value of veff may be used. Equation 9): B9 B10 B10. This gives a value for kp of 2. Middlesex.14 units.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 B8 Normally veff will not be an integer and it will be necessary to interpolate between the values given in the table. There were 5 other contributions all based on Type B evaluations for each of which infinite degrees of freedom had been assumed.7 4 4 3. based on 4 observations. TW13 4UN Website: www. U p.5 0 0 0 0 0 4 1 5. The value of tp(ν obtained from the table is the coverage factor kp that is required to calculate the ) expanded uncertainty.7 x 3 21.5 units using Equations (4) and (5).ukas. Linear interpolation will suffice for veff >3. uc(y). Unless otherwise specified. Example In a measurement system a Type A evaluation. the coverage probability p will usually be 95%. Feltham. immediately lower than 21. The expanded uncertainty is 5.5 4 4 B10. Alternatively.1 eff 5 .1 is 20. from Up = kp uc(y). 2147 High Street. UKAS Copyright exists in all UKAS publications PAGE 30 OF 82 . The combined standard uncertainty.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. gave a value of ui (y) of 3. for a coverage probability p of 95.2 The value of veff given in the tdistribution table. United Kingdom Accreditation Service. Then using the WelchSatterthwaite formula (paragraph B3.13 and this is the coverage factor that should be used to calculate the expanded uncertainty.1 3.13 = 12.7 units.45%. higherorder interpolation should be used otherwise.
In many cases this criterion will be met. where U’ is the expanded uncertainty for the remaining components. Under these circumstances. ad . but it can be the case that a single rectangular distribution may dominate over other contributions. Similar reasoning applies here. For practical purposes. As it is reasonable to assume that the arithmetic sum of these contributions would be for a coverage probability approaching 100%. If this is large compared to 3 other contributions it is less likely that the criterion above will be met. 2147 High Street.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Consequently special consideration needs to be given to the situation in which the calculated expanded uncertainty fails to meet the criterion U arithmetic sum of the limit values of all contributions.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX C DOMINANT NONGAUSSIAN TYPE B UNCERTAINTY C1.4 C1. When it is not met then the dominant contribution. A commonly encountered example is the resolution of a a digital indicating instrument. UKAS Copyright exists in all UKAS publications PAGE 31 OF 82 .3 C1. C1.5 C1. The use of a twopart expression such as this means that when both components are imported into a subsequent uncertainty budget it is likely that ad will no longer be a dominant component and a normal distribution can be assumed for the subsequent combined standard uncertainty.1 In some measurement processes there can be one component of uncertainty derived from a Type B evaluation that is dominant in magnitude compared with the other components. the second of which is dominant and is due to the uncertainty due to the resolution of the instrument being calibrated. an example is given below: The uncertainty is stated in two parts. This will involve evaluation of a coverage factor for a stated coverage probability for the convolved distributions. the statement of coverage probability associated with the expanded uncertainty will have to be modified.7 United Kingdom Accreditation Service. treated in the normal statistical manner. however.2 C1. U.6 The same situation may be encountered with other distributions associated with Type B uncertainties. Feltham. This will have a standard uncertainty of i . When the dominant component is characterised by limits within which there is a high probability of occurrence. C1. and the suggested coverage probability statement can be modified accordingly.ukas. There may be. may be greater than the arithmetic sum of the semirange of all the individual limiting values. a calculated expanded uncertainty. Middlesex. there is a degree of pessimism in following the normal recommended procedure for combination of uncertainties. should be extracted and a new value of the expanded uncertainty should be reported as U = U’ + a d . using the coverage factor of k = 2. for which a rectangular probability distribution has been assumed. the “limit” value of a normal distribution can be taken as three times its standard deviation. TW13 4UN Website: www. An example is the Ushaped distribution associated with mismatch uncertainty in RF and microwave systems. situations where a single value of uncertainty is required even if there is a Type B uncertainty that causes the distribution to be nonnormal.
97 1. which.00 2.55 0.0000 V.45 1.90 1.77.0005 V. The expanded uncertainty of the applied voltage is ± 0.20 0.ukas.82 u i y normal u i y rect 0.99 2.000100 0.0005 Probability distribution Normal Rectangular Convolved 0.70 1.91 1.84 1.96 1.71 and u i y rect the coverage factor k will be within 5% of the usual value of 2.72 1. 2147 High Street.000541 Reported result For an applied voltage of 1. The uncertainty evaluation has been carried out in accordance with UKAS requirements.99 1. The indicator can display readings in steps of 0. If it is not.001 V. the coverage factor k for a coverage probability of 95.85 0.94 u i y normal u i y rect 0.000100 0 .30 0.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.98 1.70 0.89 1.00 0.346 0.87 1.20 1.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 C1. Middlesex.00 2.75 1.45% may be obtained from the following table: u i y normal u i y rect 0.40 0.00 1.10 0.66 1. 45 1.85 1.4 times the combined standard uncertainty for the remaining components.95 1. then i normal 0.68 1. the resulting reading is 1. UKAS Copyright exists in all UKAS publications PAGE 32 OF 82 .0002 0.92 1.40 1.77 1.001 V.90 k95.79 1.80 2. A rectangular probability distribution is assumed.000306 νor i ν eff ∞ ∞ ∞ ∞ U Expanded uncertainty 0.95 1.15 0. therefore there will be a possible rounding error of ± 0.45 1. smaller.77 Divisor 2 √ 3 ci 1 1 ui(V) V 0. TW13 4UN Website: www. The resulting standard uncertainty has been multiplied by a coverage factor k = 1. for this particular convolution. The only other uncertainty of significance is due to the rounding of the indicator display. uncertainties.000 V the voltmeter reading was 1.8 If a rectangular distribution and a normal distribution are convolved.65 0.65 1. NOTE A simple test to determine whether a rectangular uncertainty is a dominant component is to check whether its standard uncertainty u y is more than 1.0002 V (k=2).10 1.00.45 k 95. United Kingdom Accreditation Service.00054V.001 V ± 0.00 C1.000289 Convolved k = 1.35 0.80 0.000289 0.50 ∞ k95 . corresponds to a coverage probability of approximately 95%. The reported expanded uncertainty is based on a convolution of a dominant rectangular uncertainty with other.75 0.93 1. The uncertainty budget will therefore be as follows: Symbol Vs δ Id uc (V) Source of uncertainty Uncertainty of applied voltage Digital rounding of indicator Combined standard uncertainty Value ±V 0.60 0.25 0.50 0. Feltham.95 1.9 Example A digital voltmeter is calibrated with an applied voltage of 1.
50 0.45 1.70 0.41 1.62 1.0 2.95 1.74 u i y normal ui y U shaped k 95.96 1.90 1. either rectangular or Ushaped.20 0.30 0.70 0.82 1.60 1.75 1.74 1.85 1.57 1.89 1.88 1.35 0.5 10 20 ∞ If two distributions of identical form.25 0.99 2.15 0.45 C1.00 0.35 0.80 0.20 1.77 1.00 United Kingdom Accreditation Service.0 7.45 1.86 0.75 1.67 1.00 1.00 2.75 0.40 0.93 1.92 1.77 1.45 0.60 0. 2147 High Street.50 0.41 1.64 1. Feltham.90 1.55 0.65 1.15 0.05 0.00 0.72 1.15 0.53 1.86 1.45 1.51 1.0 6.65 1.65 0.71 1.62 1.86 u i y rect ui y U shaped k 95. are convolved.80 1.25 0.40 C1.90 1.89 1.86 1.45 0.85 0.45 1.50 ∞ If a Ushaped distribution and a rectangular distribution are convolved.30 0.0 4.00 2. the coverage factor k for a coverage probability of 95.65 0.45% may be obtained from the following table: u i y rect ui y U shaped k 95.78 1.99 1.40 1.69 1.80 0.40 0.10 0.00 0.66 1.90 0.93 k for stated ratio 2 Ushaped Distributions 1.11 0.45 1.77 1.65 1.82 1.69 1.88 1.81 1.0 3.35 0.80 2.70 0.20 0.68 1.73 u i y rect ui y U shaped k 95.97 1.58 1.10 1.93 1.10 If a Ushaped distribution and a normal distribution are convolved.86 1.12 0.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.10 0.50 0.90 1.79 1.66 1.45% may be obtained from the following table: u i y smaller u i y larger Ratio k for stated ratio 2 Rectangular Distributions 1.00 0.91 1.80 1.0 5.69 1.45% may be obtained from the following table: u i y normal ui y U shaped k 95.41 1.53 1.10 0.84 1.84 1.30 0.45 1.55 1.93 1.47 1.25 0.95 1.86 1.92 u i y normal ui y U shaped k 95.60 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 C1.20 0.83 1.71 1. TW13 4UN Website: www.89 1.49 1.80 0. UKAS Copyright exists in all UKAS publications PAGE 33 OF 82 .66 1.75 1.66 1. the coverage factor k for a coverage probability of 95. Middlesex.ukas. the coverage factor k for a coverage probability of 95.82 1.48 1.70 1.60 0.44 1.72 1.
There are. are compared by connecting them in series and passing a constant current through them. Feltham. The mathematical model of the measurement process is used to identify the input quantities that need to be considered in the uncertainty budget and their relationship to the total uncertainty for the measurement. … xN ) For example. the standard uncertainty associated with the estimated value of each one is represented by u(xi). As the same current passes through both resistors the ratio of the two voltages VS and V X will be the same as the ratio of the two resistance values. Standard uncertainty and its evaluation are discussed in Sections 4 and 5. Where. there are several input quantities. the value of RS will change with time. In this example two resistors. therefore a contribution δ D has to be included – the model now R becomes: D1.2 V R X S δ D X R R VS United Kingdom Accreditation Service. as in most cases. the calibration uncertainty of RS means that the same relative uncertainty will exist in the value of RX.3 Example The following example of how a mathematical model can be derived relates to example K1 in Appendix K. however.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX D DERIVATION OF THE MATHEMATICAL MODEL D1 D1. TW13 4UN Website: www.I) = V/I. The measurement process can usually be modelled by a functional relationship between the values of the estimated input quantities and that of the output estimate in the form y = f (x1 . Middlesex. given the symbol x. contribute to the estimated value y of the measurand or output quantity Y. UKAS Copyright exists in all UKAS publications PAGE 34 OF 82 .1 Measurement process In general a measurement process can be regarded as having estimated input quantities Xi whose values. RX V X RS VS If the resistance of RS is known the value of RX can be determined by rearranging the equation as follows: VX R X RS VS It is known that there will be various uncertainties associated with the measurement. x2 .e.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. D1. For example. if electrical resistance R is measured in terms of voltage V and current I then the relationship is R = f(V. further uncertainties to be considered.ukas. The voltage across each is measured. RS and RX. In the expression above. 2147 High Street. i.
com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.g. D2.e.DS . Any systematic error. Middlesex. As the parameters directly observed during the calibration are the voltages V X and VS. in the voltmeter reading will cancel because it is the same for both voltages . the output quantity is obtained from only the multiplication or division of the input quantities. Taking the partial derivatives of the expression describing this curve is not a straightforward process.the ratio will be unchanged. This value for the temperature coefficient. which yield the experimental standard deviation of the mean.1 Model descriptions If the functional relationship is an addition or subtraction of the input quantities. TW13 4UN Website: www. UKAS Copyright exists in all UKAS publications PAGE 35 OF 82 . This is carried out in accordance with Equations (4) and (5).2 In paragraph 3. the same voltmeter is used to measure both V S and VX. e.ukas.δd . The temperature coefficient of resistors of this type is usually described by a parabolic curve. D2 D2.3 2 2 2 x u y c i u i i N N i 1 i 1 (10) If the functional relationship is a product or quotient. i. The general form is y 1 1 2 2 xNN where the exponents pi are known positive or negative numbers. 2147 High Street. The repeatability of the process also has to be included. R TC. All that have to be considered are secondary influences. V R X S δ D TC t X R R R Δ VS Another influence is the ratio VX /V S . s V . is multiplied by the value assigned for temperature variations.δ Ab) = WS + DS + δd + δ + Ab. The model then becomes: V R X R S δ D RTC t s V X R Δ VS This is how the model in Example K1 was derived. therefore the practical approach of estimated the temperature coefficient at the nominal temperature was taken. I C. such as resolution and linearity. This is a good example of negative correlation. In this case. Feltham. and it can be seen from examination of the uncertainty budget for example K1 that these influences have been considered separately. those expressed in % terms or in parts per million. The standard cx p x p p uncertainty will then be given by u c y y u p x i x i i 1 i N 2 (11) United Kingdom Accreditation Service. I C then all the input quantities will be directly related to the output quantity and the partial derivatives will all be unity. it is convenient to assess the repeatability in terms of observed changes in the ratio V X/VS . this can be transformed to a linear addition by the use of relative values.35 it is stated that the combined standard uncertainty is calculated using the expression u c y D2. using a linear approximation. for example W X = f (W S . or offset.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 Another consideration is the effect of temperature. Δ and the product is inserted in the model: t.
2 D3. so u P u V u I P V I 2 2 P = f(V. R) = V /R.4 United Kingdom Accreditation Service. Feltham. However. An example of the treatment of correlated contributions is shown in paragraph K6.4. e.1 D3. sensitivity coefficients may still be required to account for known relationships. and this is referred to as negative correlation. so 2 u u P 2 V uR P V R 2 2 2 2 u V uP uZ V = f(P. or by the errors in a particular instrument that is used for separate measurements in the same process.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications PAGE 36 OF 82 .4 Some examples of the use of relative uncertainties are P = f(V.3 D3. It may be the case that some input quantities are affected by the same influence quantity. such as a temperature coefficient. Knowledge concerning the possibility of correlation can often be obtained from the functional relationship between the input quantities and the output quantity but it may also be necessary to investigate the effects of correlation by making a planned series of measurements.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 D2. Z) = (P·Z) . Correlated input quantities The expressions given for the standard uncertainty of the output estimate. In such cases the input quantities are not independent of each other and the equation for obtaining the standard uncertainty of the output estimate must be modified. temperature. that is.g. the input quantities are independent of each other. 2147 High Street. The Guide should be consulted for a more detailed approach to dealing with correlation based on the calculation of correlation coefficients. so V 2P 2 Z ½ D2.ukas. In other cases measurement errors will always combine in one direction and this has to be accounted for by an increase in the combined standard uncertainty. This is referred to as positive correlation. Equations (10) and (11). The effects of correlated input quantities may serve to reduce the combined standard uncertainty. Relative uncertainties should not be used when the functional relationship is already an addition or subtraction.5 The use of relative uncertainties can often simplify the calculations and is particularly helpful when the input quantities and the uncertainties are already given in relative terms. TW13 4UN Website: www. I) = V·I. D3 D3. If positive correlation between input quantities is suspected but the degree of correlation cannot easily be determined then the most straightforward solution is to add arithmetically the standard uncertainties for these quantities to give a new standard uncertainty that is then dealt with in the usual manner in Equation (1) or (11). Middlesex. such as when an instrument is used as a comparator between a standard and an unknown. will only apply when there is no correlation between any of the input estimates.
It cannot be assumed that a drift will be linear. Checks against passive standards can establish whether compliance to specification is being maintained or whether a calibration with subsequent equipment adjustment is needed. At RF and microwave frequencies. ambient temperature can affect the performance of. must be expected to change to some extent with the passage of time. E1 E1. but should be confirmed by analysis of quality control and calibration data. Feltham. Passive devices such as standard resistors or highgrade RF and microwave attenuators may be expected to drift slowly with time. for example. and the values of reference standards. An estimate of such a drift has to be assessed on the basis of values obtained from previous calibrations. Devices that incorporate thermal sensing. resistance standards has to be sought or determined. Secular stability The performance of all instruments. advice is given in specialised technical publications and manufacturers’ application notes. UKAS Copyright exists in all UKAS publications PAGE 37 OF 82 . E2 E2. for example. The manufacturer’s specification can be a good starting point for assigning the uncertainty due to instrument drift. are all contributors to the uncertainty budget. TW13 4UN Website: www. whether measuring equipment or a reference standard. in some cases drawing a smooth curve through the chosen data points by hand can be quite satisfactory. The degree of complexity in curve fitting is a matter of judgement. Whenever a new calibration is obtained the drift characteristic will need reassessment. more detailed. United Kingdom Accreditation Service. Further. With complex electronic equipment it is not always possible to follow this procedure as changes in performance can be expected to be more random in nature over relatively long periods. Middlesex. Data can be assimilated readily if displayed in a graphical form. Variations in relative humidity can also affect the values of unsealed components. 2147 High Street. attenuators. The influence of barometric pressure on certain electrical measurement standards can also be significant. The magnitude of the drift and the random instability of an instrument. such as power sensors.2 E3 E3.1 Imported uncertainty The uncertainties assigned to the values on a calibration certificate for the calibration of an instrument. The corrections that are applied for drift are subject to uncertainty based on the scatter of data points about the drift characteristic.ukas.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX E SOME SOURCES OF ERROR AND UNCERTAINTY IN ELECTRICAL CALIBRATIONS The following is a description of the more common sources of systematic error and uncertainty (after correction) in electrical calibration work. Ambient temperature is often the most important influence and information on the temperature coefficient of.1 Environmental conditions The laboratory measurement environment can be one of the most important considerations when performing electrical calibrations. as well as other sources. and the accuracy required will determine the calibration interval. A curve fitting procedure that gives a progressively greater weight to each of the more recent calibrations can be used to allow the most probable value at the time of use to be assessed. can be affected by rapid temperature changes that can be introduced by handling or exposure to sunlight or other sources of heat. with brief comments about their nature. impedance standards that depend on mechanical dimensions for their values and other precision components.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.1 E2.
it will also be necessary to assess the additional uncertainty due to interpolation that this can introduce. TW13 4UN Website: www. Consequently the value of the quantity to be measured and/or its frequency may be different from any of the calibration points.5 digit.2 E5. or level of applied voltage being different when a device is in use from when it was calibrated. The error introduced by this process will be from . or there is additional frequency calibration data from other models of the same instrument. A quantisation error of ± 0. The presence of electrical noise causing fluctuations in instrument readings will commonly determine the usable resolution. referred to as resolution or “digital rounding error”. It should also be ensured that all equipment is operating within the manufacturer's stated range of supply voltages.3 E5. UKAS Copyright exists in all UKAS publications PAGE 38 OF 82 . resistive voltage dividers and attenuators at any frequency are examples of devices being affected by selfheating and/or applied voltage. however it is possible to make a good estimate of the mean position of a fluctuating pointer by eye. Similarly. E5 E5.5 digit. The effects of harmonics and noise on ac calibration signals may have an influence on the apparent value of these signals. a directgating frequency counter has a digital rounding error of ± 1 digit. harmonic distortion.2 It is also necessary to be aware of the possible effects of electrical operating conditions. scale nonlinearity. such as power dissipation. E3. the resulting error is assumed to be zero with limits of ± 0. Feltham. Middlesex. If the measurement frequency falls between two calibration frequencies. due to the random relationship between the signal being measured and the internal clock.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. the last digit actually recorded will always be subject to an uncertainty of at least ± 0.4 United Kingdom Accreditation Service. When the value of the quantity lies between two calibration values. Resolution The limit to the ability of an instrument to indicate small changes in the quantity being measured. In either case. Interpolation of calibration data When an instrument with a broad range of measurement capabilities is calibrated there are practical and economic factors that limit the number of calibration points.1 E4. 2147 High Street. Many instruments with a digital display use an analogueto digital converter (ADC) to convert the analogue signal under investigation into a form that can be displayed in terms of numeric digits. although not necessarily a property of the display itself.ukas.5 digit may not apply in all instances and an understanding of instrument operation is needed if the assigned uncertainty is to be realistic. Some instruments may also display hysteresis that.1 E5. the effects of any commonmode signals present in a measurement system may have to be accounted for.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 E3.2 (a) (b) the performance of the actual instrument being used has been explored with a swept frequency measurement system to verify the absence of resonance effects or aberrations due to manufacturing or other performance limitations. for example. is treated as a systematic component of uncertainty.5 digit (else the last digit would be one higher).3 E4 E4. As there is no way of knowing where within this range the underlying value is.5 digit is therefore present. In an analogue instrument the effect of resolution is determined by the practical ability to read the position of a pointer on a scale. For example. This “digital rounding error” of ± 0. and wherever reasonable.0. may result in further uncertainties amounting to several digits. The last displayed digit will be a rounded representation of the underlying analogue signal. consideration needs to be given to systematic errors that arise from.5 digit (else the last digit would be one lower) to + 0. Resistance standards. One can only proceed with confidence if: a theory of instrument operation is known from which one can predict a frequency characteristic.
The main effects are leakage currents to earth. They can be minimised by design of connections that are thermally symmetrical.1 United Kingdom Accreditation Service. Thus a different arrangement between calibration and subsequent use of an instrument may be the source of systematic errors. so that the Seebeck voltage in one lead is cancelled by an identical and opposite voltage in the other. The impedance and finite electrical length of connecting leads or cables may also result in systematic errors in voltage measurements at any frequency. Thermoelectric voltages If an electrical conductor passes through a temperature gradient then a potential difference will be generated across that gradient. This is known as the Seebeck effect and these unwanted.ukas.3 E8 E8. TW13 4UN Website: www. e. Corrections may be possible if impedances are known.g. UKAS Copyright exists in all UKAS publications PAGE 39 OF 82 . In some situations. In particular.1 Apparatus layout The physical layout of one item of equipment with respect to another and the relationship of these items to the earth plane can be important in some measurements.2 E7. the resulting resonance may cause the output voltage to increase with respect to its opencircuit value.1 E8. oscilloscopes and other voltage sensing instruments may so load the circuit to which they are connected as to cause significant systematic errors. Middlesex. Generally an allowance has to be made as a Type B component of uncertainty for the presence of thermal emfs. RF mismatch errors and uncertainty At RF and microwave frequencies the mismatch of components to the characteristic impedance of the measurement system transmission line can be one of the most important sources of error and of the systematic component of uncertainty in power and attenuation measurements. Similarly. it should be noted that some multifunction calibrators can exhibit a slightly inductive output impedance.3 E9 E9. In inductance measurements it is necessary to define connecting lead configuration and be conscious of the possible effects of an earth plane or adjacent ferromagnetic material. the polarity of the dc supply is reversed and an arithmetic mean is taken of two sets of dc measurements.1 E7. and electromagnetic leakage fields. The effect of mutual heating between apparatus may also need to be considered. for inductance measurements the capacitance between connecting leads may be important. interference loop currents.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 E6 E6.2 E8.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. For capacitance measurements. the inductive properties of the connecting leads may be important. parasitic voltages can cause errors in some measurement systems – in particular. ac/dc transfer measurements. 2147 High Street. where small dc voltages are being measured. particularly at higher values of capacitance and/or frequency. This means that when a capacitive load is applied. The use of fourterminal connections minimises such errors in some dc and ac measurements. Loading and cable impedance The finite input impedance of voltmeters. This is because the phases of voltage reflection coefficients are not usually known and hence corrections cannot be applied. E7 E7. Feltham.
ukas. if only the magnitude and not the phase of the directivity component is known.5 s 11a 2 2 s11b L 2 s 22 a 2 2 s22 b G 2 L 2 s 21a 4 4 s 21b 0 . Harris and Warner have shown that the standard deviation of mismatch.5 E(5) where Γ and Γ are the source and load voltage reflection coefficients respectively and s11. M. Harris and Warner concluded that the distribution for M would approximate to that of a normal distribution due to the combination of its component distributions. The uncertainty will be equal to the directivity. Middlesex.2 In a power measurement system. while the limits of L mismatch uncertainty are ± 2 Γ Γ .g.3 2 2 1 G cos L E(2) When φis unknown.414 E(4) G L 2 When a measurement is made of the attenuation of a twoport component inserted between a generator [3] and load that are not perfectly matched to the transmission line. UKAS Copyright exists in all UKAS publications PAGE 40 OF 82 .02 2 . PO. expressed in dB is approximated by 8. s22.Γ and is accounted for in the calibration factor. if the measured value of Γ is 0.03 ± 0. Harris and Warner show that the distribution is Ushaped with a standard deviation given by 2 G L u(mismatch) = 1. this expression for absorbed power can have limits: PL P 0 (limits) = 1 2 G L 2 1 L 2 E(3) E9.036. this becomes G L PL P0 1 L E9. United Kingdom Accreditation Service.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. e. P L. expressed in linear terms. the finite directivity of the bridge or reflectometer gives rise to an uncertainty in the measured value of the VRC. TW13 4UN Website: www. that would be absorbed in a load equal to the [3] characteristic impedance of the transmission line has been shown to be related to the actual power. E9. When G L Γ and Γ are small. a directivity of 30 dB is equivalent to an uncertainty of ± 0. i.02 then the value of Γ that should be used to calculate L L the mismatch uncertainty is E10 E10. for example. 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 E9.6 The values of Γ and Γ used in Equations E(4) and E(5) and the scattering coefficients used in Equation G L E(5) will themselves be subject to uncertainty because they are derived from measurements. Feltham.e. absorbed in a wattmeter terminating the line by the equation PL 2 cos 2 2 P0 1 G L G L E(1) 2 1 L where φis the relative phase of the generator and load voltage reflection coefficients Γ and Γ. This uncertainty has to be considered when calculating the mismatch uncertainty and it is recommended that this is done by adding it in quadrature with the measured or derived value of the reflection coefficient. 686 M G 2 2 E9. the power.03 2 0.0316 VRC. Because a cosine function characterises the probability L G [3] distribution for the uncertainty.1 Directivity When making voltage reflection coefficient (VRC) measurements at rf and microwave frequencies. 0. 2147 High Street.4 The calculable mismatch error is 1 . s21 are the G L scattering coefficients of the twoport component with the suffix a referring to the starting value of the [3] attenuator and b referring to the finishing value of the attenuator.
is calculated from u(TP) = 2 TP. for example.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 E10. expressed as a VRC. TW13 4UN Website: www. Although connecting and disconnecting the device can evaluate the repeatability of particular connector pairs in use. Middlesex. Reference [5] provides advice on the specifications and use of coaxial connectors including guidance on the repeatability of the insertion loss of connector pairs. E12 United Kingdom Accreditation Service.012 = 0. u(TP). these connector pairs are only samples from a whole population. X X When a directional coupler is used to monitor incident power in the calibration of a power meter it is the effective source match of the coupler that defines the value of Γ referred to in E9. and Γ is the measured reflection coefficient. UKAS Copyright exists in all UKAS publications PAGE 41 OF 82 .4 dB). Feltham. RF connector repeatability The lack of repeatability of coaxial pair insertion loss and. 0 E11 E11. to a lesser extent.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.6 above it is recommended that the uncertainty in the measurement of directivity is taken into account by adding the measured value in quadrature with the uncertainty.ukas. in linear quantities. As with E9. G the measured value of test port match will have an uncertainty that should be taken into account by using quadrature summation. if the measured directivity of a bridge is 36 dB (0.01) then the directivity to be used is 0.1 Test port match The test port match of a bridge or reflectometer used for reflection coefficient measurements will give rise to an error in the measured VRC due to rereflection.016 2 . voltage reflection coefficient is a problem when calibrating devices in a coaxial line measurement system and subsequently using them in some other system. The uncertainty.6 and E10.016) and has an uncertainty of +8 dB 4 dB (± 0. 2147 High Street.019 (34. To obtain representative data for guidance for various types of connectors in use is beyond the resources of most measurement laboratories.1 As with E9.Γ where TP is the test port match.
If such a history is not available. the length of arm error (assuming it is constant) of an equal arm balance need not be assessed if the weighing process only uses substitution techniques (Borda's method). eccentricity (off centre load). surface finish. then it is usual to assume that they may change in mass by an amount equal to their uncertainty of calibration between calibrations. especially if groups of weights are placed on the weighing pan simultaneously. Weighing machine/weighing process The performance of the weighing machine used for the calibration should be assessed to estimate the contribution it makes to the overall uncertainty of the weighing process. illfitting screw knobs).g. Feltham. length of arm error. i. Middlesex.1 F2 F2. magnetic effects (e.1 (a) (b) (c) (d) (e) (f) (f) United Kingdom Accreditation Service. differences between the temperature of the weights and the weighing machine. For example. The performance assessment should cover those attributes of the weighing machine that are significant to the weighing process. TW13 4UN Website: www. The calibration interval for reference weights will depend on the stability of the weights. Further information about mass calibration can be found in reference [6]. readability.g. UKAS Copyright exists in all UKAS publications PAGE 42 OF 82 . physical wear and damage and atmospheric contamination. but their effect should at least be considered when estimating the overall uncertainty of a measurement. The stability of weights can be affected by the material and quality of manufacture (e. Secular stability of reference weights It is necessary to take into account the likely change in mass of the reference weights since their last calibration. The assessment may include some or all of the following: repeatability of measurement. The figure adopted for stability will need to be reconsidered if the usage or environment of the weights changes. digit size/weight value per division. or the effect of force balance motors on cast iron weights). unstable adjustment material.ukas. F1 F1. 2147 High Street.e. linearity within the range used.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. They may not all be significant at all levels of measurement.g. F3 F3. This change can be estimated from the results of successive calibrations of the reference weights. e.1 Reference weight calibration The uncertainties assigned to the values on a calibration certificate for the calibration of the reference weights are all contributors to the uncertainty budget. temperature effects. magnetic weights.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX F SOME SOURCES OF ERROR AND UNCERTAINTY IN MASS CALIBRATIONS This Appendix describes the more common sources of errors and uncertainties in mass calibration with brief comments about their nature..
Feltham. UKAS Copyright exists in all UKAS publications PAGE 43 OF 82 . German silver or gunmetal weights this effect may be up to ±16 6 6 parts in 10 .1 Air buoyancy effects The accuracy with which air buoyancy corrections can be made depends on how well the density of the weights is known. For cast iron weights the figure may be up to ±18 parts in 10 and for aluminium weights up to 6 ± 45 parts in 10 . For mass comparisons a figure of ± 1 part in 10 of the applied mass is typical for common weight materials such as stainless steel. as well as compensating for changes in the machine itself. pressure and humidity. For the highest levels of accuracy. Apart from air buoyancy effects.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 F4 F4.079 kg m to 1. the environment in which the calibration takes place can introduce uncertainties. 6 F4. 2147 High Street.2 F4. The movement of weights during the calibration causes disturbances to the local environment. Changes in the level of humidity in the laboratory can make shortterm changes to the mass of weights. Rapid changes of temperature in the laboratory can affect the weighing process. as will draughts from air conditioning units. The air density is usually calculated from an equation (see reference [6]) after measuring the air temperature. and ± 125 parts in 10 and ± 155 parts in 10 respectively for direct weighing.291 kg m which can be produced by ranges of relative humidity from 30% to 70%. German silver and gunmetal. while low levels of humidity can introduce static electricity effects on some comparators.4 F4. It is common practice to reduce the effects of buoyancy on such devices by the use of an auxiliary weight. This spanning weight can be external or internal to the machine. TW13 4UN Website: www.ukas. Middlesex. Some laboratories can determine the density of weights. but for most mass work assumed figures are used. known as a spanning weight. For the ambient conditions stated above the uncertainty limits due to buoyancy effects may be ± 110 parts 6 6 in 10 and ± 140 parts in 10 respectively for comparing water and organic solvents with stainless steel 6 6 mass standards. leaving just the uncertainty of the correction. and how well the air density can be determined. If such machines are not spanned at the time of use the calibration may be subject to an increased uncertainty due to the buoyancy effects on the loading weights. which will affect the reading. The 3 3 figures that follow are based upon an air density range of 1. plated brass. air temperature from 10 to 30 and C C barometric pressure from 950 millibar to 1050 millibar. pressure and humidity. Certain weighing machines display mass units directly from the force they experience when weights are applied. The uncertainty can be reduced if the mass comparisons are made within suitably restricted ranges of air temperature. it may also be necessary to measure the carbon dioxide content of the air.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Temperature gradients can give rise to convection currents in the balance case. Dust contamination also introduces errors in calibrations. plated brass.5 United Kingdom Accreditation Service. If corrections are made for the buoyancy effects the uncertainty can be virtually eliminated. which is used to normalise the readings to the prevailing conditions.3 F4. For cast iron the figure may 6 6 be up to ± 3 parts in 10 and for aluminium up to ± 30 parts in 10 . For weighing machines that make use of stainless steel.
e. Similarly any thermal emfs introduced by switches or scanner units should be investigated. When partial immersion liquidinglass thermometers are to be calibrated an additional uncertainty contribution to account for effects arising from differences in depth of immersion should be included even when the emergent column temperature is measured.. When platinum resistance thermometers are used as reference standards any contribution to the uncertainty due to selfheating effects should be considered. in applying scale corrections or deviations from a reference table. the calibration bath or furnace. including any contribution due to difference in immersion of the reference standard from that stated on its certificate of calibration. Mathematical interpretation Uncertainty arising from mathematical interpretation. should be assessed.1 United Kingdom Accreditation Service.1 G4. Measuring instruments The uncertainty assigned to the calibration of any electrical or other instruments used in the measurements. TW13 4UN Website: www. Instability and temperature gradients in the thermal environment. In such cases. Resolution of reading. e.1 (a) (b) Reference thermometer calibration The uncertainties assigned to the values on a calibration certificate for the calibration of the reference thermometer are all contributors to the uncertainty budget. For many calibrations.g. possibly by varying the immersion depth of the thermocouple in an isothermal enclosure. Each source may have several uncertainty components.ukas. Ideally this should be evaluated at the time of calibration.1 G3 G3.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX G SOME SOURCES OF ERROR AND UNCERTAINTY IN TEMPERATURE CALIBRATIONS The more common sources of systematic error and uncertainty in the measurement of temperature are described in this section. this will not be practical. 2147 High Street.. `in air' or in stirred liquid. (c) (d) G4 G4. standard resistors.1 G2 G2.g. When thermocouples are being calibrated any uncertainty introduced by compensating leads and reference junctions should be taken into account.g. UKAS Copyright exists in all UKAS publications PAGE 44 OF 82 . this may be very significant in the case of a liquidinglass thermometer or digital thermometers. a figure of 20% of the maximum permissible error for the particular thermocouple type is considered reasonable. however. Middlesex.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Feltham. Further influences Additional uncertainties in the measurement of the temperature using the reference thermometers: Drift since the last calibration of the reference thermometers and any associated measuring instruments. e.2 G5 G5. G1 G1.g. measuring bridges and digital multimeters. or in curvefitting to allow for scale nonlinearity. Unknown errors arising from inhomogeneity of the thermocouple being calibrated can give rise to significant uncertainties. Contributions associated with the thermometer to be calibrated These may include factors associated with electrical indicators as well as some of the further influences already mentioned. This will mainly apply if the measuring current is different from that used in the original calibration and/or the conditions of measurement e.
Whilst mathematical corrections can be made there will be residual uncertainties resulting from the uncertainty of the measuring force and in the properties of the materials involved. Small residual errors can still result where.1 H6 H6. Secular stability of reference standards and instrumentation The changes that occur over time must be taken into account. Typically these will include small errors in the flatness or sphericity of stylus tips. This is particularly important when the equipment may be exposed to physical wear as part of normal operation. 2147 High Street.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.ukas. Geometric errors Errors in the geometry of the gauge being calibrated. flatness. the straightness.1 H5 H5. TW13 4UN Website: www.1 H4 H4. with respect to the axis of measurement. Feltham. Cosine errors Any misalignment of the gauge being calibrated or reference standards used. any reference standards used or critical features of the measuring instruments used to make the measurements can introduce additional uncertainties. They are likely to be most significant in the more precise calibrations and in cases involving dissimilar materials. Whilst it may be possible to make corrections for temperature effects there will be residual uncertainties resulting from uncertainty in the values used for the coefficients of expansion and the calibration of the measuring thermometer. Such errors are often referred to as cosine errors and can be minimised by adjusting the attitude of the gauge with respect to the axis of measurement to find the relevant turning points that give the appropriate maximum or minimum result. Temperature effects The uncertainties associated with differences in temperature between the gauge being calibrated and the reference standards and measuring instruments used should be accounted for. usually by reference to the calibration history of the equipment. for instance. parallelism or squareness of surfaces used as datum features. Such errors are often most significant in cases where perfect geometry has been wrongly assumed and where the measurement methods chosen do not capture. H1 H1. H3 H3. incorrect assumptions are made concerning any features used for alignment of the datums.1 Reference standards and Instrumentation The uncertainties assigned to the reference standards and those for the measuring instruments used to make the measurements. will introduce errors into the measurements.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX H SOME SOURCES OF ERROR AND UNCERTAINTY IN DIMENSIONAL CALIBRATIONS The more common sources of systematic error and uncertainty in dimensional measurements are described in this section. Middlesex. UKAS Copyright exists in all UKAS publications PAGE 45 OF 82 . Elastic compression These are uncertainties associated with differences in elastic compression between the materials from which the gauge being calibrated and the reference standards were manufactured. These will be most significant over the longer lengths and in cases involving dissimilar materials. They will relate to the measuring force used and the nature of stylus contact with the gauge and reference standard. suppress or otherwise accommodate the geometric errors that prevail in a particular case. and the roundness or taper in cylindrical gauges and reference standards.1 United Kingdom Accreditation Service.1 H2 H2.
With care. and material composition and is therefore a best estimate from actual data. J2. Feltham. J1 J1. The secular stability uncertainty for the area will depend on the calibration interval and can be larger than the calibration uncertainty. Piston and weight carrier mass. Reference DWT mass set uncertainty The uncertainties assigned to the values on a calibration certificate for the weights in the reference dead weight tester mass set are all contributors to the uncertainty budget.2 J3 J3. Where this is not available it is recommended that a pessimistic estimate is made and a short calibration interval set. TW13 4UN Website: www. Some knowledge of the Bouguer anomalies is required to achieve these levels of uncertainty from such calculations.ukas.1 J4 J4. The drift of the piston mass will be larger in oil DWTs as this will reflect the difficulties in repeat weighting of pistons that have been immersed in oil. Middlesex. 2147 High Street. using dead weight testers (DWT). this can be measured with an uncertainty of less than 1 ppm.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications PAGE 46 OF 82 . The variation between calibrations in the area of a DWT will depend on its usage. This change can be estimated from successive calibrations of the reference DWT.1 United Kingdom Accreditation Service. The relative uncertainty is often higher at lower pressures. This uncertainty will often vary with pressure. It is possible for an estimate of the g value to be obtained from a reputable geological survey organisation based on a grid reference. Secular stability of the reference DWT mass set It is necessary to account for likely changes the mass set of the reference DWT since the last calibration. this would attract an uncertainty of around 3 ppm. It can also be calculated from knowledge of latitude and altitude.1 J5 J5. These difficulties arise from incomplete cleaning processes and possible instability due to the evaporation of solvents.1 Reference DWT The uncertainties assigned to the values on a calibration certificate for the reference dead weight tester are all contributors to the uncertainty budget. g.1 addresses the subject of secular stability of reference weights.around 50 ppm in the UK. design. These include the following: Area uncertainty including any uncertainty in the distortion. Paragraph F2. Secular stability of the reference dead weight tester It is necessary to account for likely changes in the area and mass of the reference DWT since the last calibration. however the uncertainty will be much larger .THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX J SOME SOURCES OF ERROR AND UNCERTAINTY IN PRESSURE CALIBRATIONS USING DEAD WEIGHT TESTERS The more common sources of systematic error and uncertainty in the generation of known pressures. are described in this section. Uncertainty of local gravity determination The pressure generated by a DWT is directly affected by the local acceleration due to gravity.1 (a) (b) J2 J2. The uncertainty of the mass stack used to generate pressure should be evaluated over the range of the DWT.
1 J10 J11 J11. its temperature coefficient of expansion being related to the particular materials that the piston and cylinder are made from.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 J6 J6.ukas. variation in temperature during a calibration and uncertainty in the determination of the piston temperature. For hydraulic calibrations the effect is a fixed pressure effect that will depend on the density of the fluid used. then all of the force would act on the area.h) For most DWT oils the effect is between 8 and 9 Pa/mm. buoyancy volume corrections and surface tension corrections will also need to be considered.1 Air buoyancy effect Air buoyancy affects the mass set of a DWT in the same way as described in paragraph F4. This effect will be related to the fall rate of the piston and the particular measurement procedure in use. (Fluid head pressure = ρ . Nonverticality of the DWT piston An uncertainty arises due to the fact that the piston may not be perfectly vertical. These figures are usually reported on calibration certificates for DWTs. 2147 High Street. care must be taken to convert any quoted correction to the actual oil used if different from that used during the calibration of the reference DWT.1 Uncertainties arising from the calibration process Any uncertainty arising from the calibration process will need to be evaluated. Uncertainty due to head correction Any difference between the height of the reference DWT datum level and that of the item being calibrated will affect the pressure generated at that item.1 J7 J7.g. The effect in terms of generated pressure is proportional to the cosine of the angle from true vertical. United Kingdom Accreditation Service. These could include the resolution and repeatability of the unit being calibrated and the effects of the environment on it. TW13 4UN Website: www. UKAS Copyright exists in all UKAS publications PAGE 47 OF 82 . local acceleration due to gravity and the height difference. Middlesex.1 J8. Uncertainties due to calculation or data fitting of the calibration results may also have to be considered.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.2 J9 J9. Effects of fluid properties For hydraulic calibrations the effect of the fluid properties on fluid head corrections. Temperature effect on DWT area The area on a DWT changes with temperature. In most circumstances the uncertainty of these influence quantities can be treated as negligible. However. J8 J8. If it were. Any departure from vertical will reduce the force and therefore the generated pressure. The float height position of the piston will also contribute to the head correction uncertainty. Consideration has to be given to any variation in temperature from the reference temperature when the DWT was calibrated. Feltham. For pneumatic calibrations this effect is proportional to pressure and normally equates to about 116 ppm/m.
3 = = = = = = = Calibration value for the standard resistor. Middlesex. was estimated to have limits of ± 0. The value of the unknown resistor.5 ppm at a coverage probability of approximately 95% (k = 2).ukas. These examples may also be used for the purpose of software validation. They are presented to illustrate the principles involved in uncertainty evaluation and to show how the common sources of uncertainty in the various fields can be analysed in practice.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. The temperature coefficient of resistance for the standard resistor was obtained from a graph of temperature versus resistance.2 ppm for each reading. Relative temperature coefficient of resistance for R S.5 ppm. The examples are not intended as preferred or mandatory requirements. Voltage across RS .5 K1. The calibration certificate for the standard resistor reported an uncertainty of ± 0.4 K1. The uncertainty in this correction. 2147 High Street. Calibration of a 10 kΩ resistor by voltage intercomparison A highresolution digital voltmeter is used to measure the voltages developed across a standard resistor and an unknown resistor of the same nominal value as the standard. RD. the configuration of the spreadsheet can be verified by entering the same values and comparing the output of the spreadsheet with the results shown in the examples. Each of these is assigned a rectangular distribution. which was estimated to have limits of ±0. Records of evaluation of the oil bath characteristics showed that the maximum temperature deviation from the set point did not exceed ± 0.1ºC at any point within the bath. TW13 4UN Website: www. Voltage across RX .5 ppm per ºC was assigned. however. They are. A correction was made for the estimated drift in the value of RS. believed to be realistic for the particular measurements described. Maximum variation in oil bath from nominal temperature.1 RS δD R RTC Δ t VX VS sV K1. however using a linear approximation over the small range of temperature variation encountered in the bath.0°C. K1. when the seriesconnected resistors are supplied from a constant current dc source. This value was included in the uncertainty budget as a sensitivity coefficient. RX . Such curves are normally parabolic in nature. If an uncertainty budget has been prepared using a spreadsheet. Feltham. Repeatability of ratio V X/VS. a value of ± 2. is given by V R X RS R D TC t s V X .THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX K EXAMPLES OF APPLICATION FOR CALIBRATION NOTES (a) This Appendix presents a number of example uncertainty budgets in various fields of measurement. where δ R Δ VS (b) K1 K1.2 K1. UKAS Copyright exists in all UKAS publications PAGE 48 OF 82 .6 United Kingdom Accreditation Service. Both resistors are immersed in a temperature controlled oil bath maintained at 20. Relative drift in R S since the previous calibration. The same voltmeter is used to measure VX and V S and although the uncertainty contributions will be correlated the effect is to reduce the uncertainty and it is only necessary to consider the relative difference in the voltmeter readings due to linearity and resolution.
5 From Equation (3). 2147 High Street.5 ppm.ukas. Feltham.8 Uncertainty budget Value ± 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K1. UKAS Copyright exists in all UKAS publications PAGE 49 OF 82 . 5 K1.89 ppm The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2.115 0.5 ppm 0.9 Reported result The measured value of the 10 kΩresistor at 20 ºC ± 0. +10. u s V V = 0.115 0.4. Middlesex. +10.25 0. providing a coverage probability of approximately 95%.445 0. +10. the mean value V = +10. 0 .7.105 Ω ± 0.289 0. monitored source of defined source impedance. The relevant temperature conditions are therefore included in the reporting of the result.3.071 ppm Probability distribution Normal Rectangular Rectangular Rectangular Rectangular Normal Normal Normal (k = 2) ui(RX) ppm 0.144 0.7 Type A evaluation: Five measurements were made to record the departure from unity in the ratio V X /V S in ppm. TW13 4UN Website: www. K2 K2.6.5 ppm/ºC 1 1 1 ∞ ∞ ∞ ∞ ∞ 4 >500 >500 K1.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.5 ppm 0.158 From Equations (5) and (6). NOTE The temperature coefficient of the resistor being calibrated is normally not included.1 Calibration of a coaxial power sensor at a frequency of 18 GHz The measurement involves the calibration of an unknown power sensor against a standard power sensor by substitution on a stable. +10.0707 ppm.071 0. The uncertainty evaluation has been carried out in accordance with UKAS requirements.2 ppm 0.2 ppm 0. The readings were as follows: +10. defined as . for the same Incident power at calibration frequency power sensor response and is determined from the following: United Kingdom Accreditation Service.1 ºC 0.1º C was 10 000. The measurement is Incident power at reference frequency made in terms of Calibration Factor.891 νor i ν eff Symbol RS δD R Δ t VS VX u(V) uc(y) U Source of uncertainty Calibration of standard resistor Uncorrected drift since last calibration Temperature effects Voltmeter across R S Voltmeter across R X Repeatability of indication Combined standard uncertainty Expanded uncertainty Divisor 2 √ 3 √ 3 √ 3 √ 3 1 ci 1 1 2. as it is an unknown quantity.
the mean value K X = 93.02% From Equation (3).12 at 18 GHz. K2. S Γ = 0. The error due to secular stability was therefore assumed to be zero with limits. A value of ± 0.4% per year.20% 93. G Γ = 0.7415 From Equations (5) and (6).02 at 50 MHz and 0. K2. Ratio of Mismatch Losses. TW13 4UN Website: www. X These values include the uncertainty in the measurement of Γas described in paragraph E9.1% was quoted for a coverage factor k = 2. Four separate measurements were made which involved disconnection and reconnection of both the unknown sensor and the standard sensor on a power transfer system. These will be 200ΓΓ% and 200ΓΓ% respectively. Feltham.10 at 18 GHz. The instrumentation linearity uncertainty was estimated from measurements against a reference attenuation standard.7 K2.5 The standard power sensor was calibrated by an accredited laboratory 6 months before use. 2147 High Street.45% 92. The longterm stability of the standard sensor was estimated from the results of 5 annual calibrations. All measurements were made in terms of voltage ratios that are proportional to calibration factor. Drift in standard sensor since the previous calibration. Type A evaluation: The four measurements resulted in the following values of Calibration Factor: 93. The expanded uncertainty for k = 2 of ± 0.6 K2. Middlesex. Ratio of reference power source (shortterm stability of 50 MHz reference). u R s K X K = 0. not greater than ± 0. There will be mismatch uncertainties associated with the source/standard sensor combination and with the source/unknown sensor combination. where KS DS δ DC δ M δ REF K2.02 at 50 MHz and 0.2% was used as the previous calibration was within 6 months.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.3707%. in this case.ukas. Ratio of DC voltage outputs.95% 93.6.2 = = = = = Calibration Factor of the standard sensor.02 at 50 MHz and 0. No predictable trend could be detected so drift corrections could not be made.4 K2. where G S G X Γ = 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 Calibration Factor. UKAS Copyright exists in all UKAS publications PAGE 50 OF 82 .1% applies to ratios up to 2:1.07 at 18 GHz. the expanded uncertainty of ± 1. KX = (KS + DS) x δ x δ x δ DC M REF. 4 United Kingdom Accreditation Service. None of the uncertainty contributions are considered to be correlated to any significant extent.16%.8 0 .3 K2.
0 1. Drift in reference IF attenuator since last calibration. within the timescale allowed for its shortterm stability.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Mismatch error.0 1. Error due to resolution of detection system. This example illustrates the significance of mismatch uncertainty in measurements at relatively high frequencies.37 Probability distribution Normal Rectangular Normal Rectangular Ushaped Ushaped Ushaped Ushaped Normal Normal Normal (k = 2) ui(KX) % 0.1 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K2. 2147 High Street. The uncertainty evaluation has been carried out in accordance with UKAS requirements. Effect of signal leakage.37 1. the uncertainty in the absolute value of the 50 MHz reference source need not be included if the standard and unknown sensors are calibrated using the same source.0 ci 1. Repeatability. 2 3 K3 K3.116 0. where Ab Aa A IF DIF LM RD M AL AR = = = = = = = = = Indicated attenuation with unknown attenuator set to zero. The measurement is made in terms of the attenuation in dB between a matched source and load from the following: A X = A b – Aa + AIF + DIF + L M + R D + M + A L + AR .116 0. The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2.0 ∞ ∞ ∞ ∞ ∞ ∞ ∞ ∞ 3 >500 >500 K2.05 0.99 1. NOTES 1 For the measurement of calibration factor. Indicated attenuation with unknown attenuator set to 30 dB. UKAS Copyright exists in all UKAS publications PAGE 51 OF 82 .69 3. United Kingdom Accreditation Service.55 0.ukas.0 1.08 1.08 0. TW13 4UN Website: www.0 1.10 Reported result The measured calibration factor at 18 GHz was 93. In a subsequent use of a sensor further random components of uncertainty may arise due to the use of different connector pairs.0 1.19 0.4 %. Calibration of reference IF attenuator.06 0.0 1. providing a coverage probability of approximately 95%.39 νor i ν eff Symbol KS DS δ DC δ M M1 M2 M3 M4 KR u(KX) U Source of uncertainty Calibration factor of standard Drift since last calibration Instrumentation linearity Stability of 50 MHz reference Mismatch: Standard sensor at 50 MHz Unknown sensor at 50 MHz Standard sensor at 18 GHz Unknown sensor at 18 GHz Repeatability of indication Combined standard uncertainty Expanded uncertainty Divisor 2. Feltham.9 Uncertainty budget Value ±% 1.2 0.1 Calibration of a 30 dB coaxial attenuator The measurement involves the calibration of a coaxial step attenuator at a frequency of 10 GHz using a dual channel 30 MHz IF substitution measurement system.0 √ 3 √ 2 √ 2 √ 2 √ 2 1.0 1.2 % ± 3. Departure from linearity of mixer.40 1.06 0.1 0. Middlesex.0 √ 3 2.0 1.68 0.2 0.
The probability distribution is assumed to be rectangular. which gave an uncertainty of ± 0.9 0 .002 0.018 From Equations (5) and (6).0491 νor i ν eff Symbol Source of uncertainty Calibration of reference attenuator Drift since last calibration Mixer nonlinearity Resolution of indication Mismatch Signal leakage effects Repeatability of indication Combined standard uncertainty Expanded uncertainty Divisor 2.0016 0.005 dB.0 ci 1.0025 0.e.06 dB From Equation (3).031 K3.0041 0. Middlesex. The probability distribution is assumed to be rectangular.4 K3.001 dB could be observed for attenuation values up to 70 dB. No correction is made for the drift of the IF attenuator. K3.ukas.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K3.05 s22b = 0. No correction is made for mixer nonlinearity.006 dB was therefore assigned at 30 dB.01 dB with a triangular distribution.01 s21a = 1 s21b = 0. The limits of ± 0. Two identical rectangular distributions with semirange limits of a combine to give a triangular distribution with semirange limits of 2a.005 0.3 K3.09 s22a = 0. The mismatch uncertainty is calculated from the scattering coefficients using Equation E(5).0 1.009 0.006 0.009 dB.5 K3. The results were as follows: 30.022 0. No effect greater than ± 0.05 s11b = 0.005 dB at a coverage probability of 95% (k = 2).050 dB.03 L Γ = 0. An uncertainty of ± 0. The values used were as follows: Γ = 0.0 1.009 Probability distribution Normal Rectangular Rectangular Triangular Normal Rectangular Normal Normal Normal (k = 2) ui(AX) dB 0. Feltham.6 K3. Type A evaluation: Four measurements were made which involved setting the reference level with the step attenuator set to zero and then measuring the attenuation for the 30 dB setting.0 1. the mean value A X = 30.03 G s11a = 0.7 Special experiments were performed to determine whether signal leakage had any significant effect on the measurement system.once for the 0 dB reference setting and again for the 30 dB measurement.03 dB 30. TW13 4UN Website: www. ± 0.04 dB 30.002 dB/10dB.022 0. This occurs twice . UKAS Copyright exists in all UKAS publications PAGE 52 OF 82 .8 K3.0 √ 3 √ 3 √ 6 1 √ 3 1. The uncertainty was estimated from a series of linearity measurements over the dynamic range of the system to be ± 0.07 dB 30. u R s A X A = 0.2 The result is corrected for the calibrated value of the IF attenuator using the results from a calibration certificate.0 A IF DIF LM RD M AL AR u(AX) U ∞ ∞ ∞ ∞ ∞ ∞ 3 >500 >500 United Kingdom Accreditation Service. 2147 High Street.0 1.010 0.001 0.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. The resolution of the detection system was estimated to cause possible rounding errors of onehalf of one least significant recorded digit i.0 1.002 dB were estimated from the results of previous calibrations.0035 0. No correction is made for mismatch error.0006 0.0 1. 4 Uncertainty budget Value ± dB 0. The uncertainty due to resolution is therefore 0.0246 0.
comprising 10 comparisons between standard and unknown. The linearity error of the comparator over the 2. gave a standard deviation. and therefore includes effects due to eccentricity errors.1 Calibration of a weight of nominal value 10 kg of OIML Class M1 The calibration is carried out using a mass comparator whose performance characteristics have previously been determined. ± 10 mg.3 K4.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Thus relatively small uncertainties expressed in dB may be combined in the same way as those expressed as linear relative values. In a subsequent use of an attenuator further random components of uncertainty may arise due to the use of different connector pairs. 2147 High Street.6 K4.5. that is ± 10 mg. x. Combining these two rectangular distributions gives a triangular distribution.5 K4.g.023 and 2.e. Middlesex. 2 For attenuation measurements. The unknown weight is obtained from: Measured value of unknown weight.1 dB corresponds to a power ratio of 1.4 K4. Digital rounding δd has limits of I ± 0.2 The calibration certificate for the standard mass gives an uncertainty of ± 30 mg at a coverage probability of approximately 95% (k = 2).023) = 0. The rounding of the value of the least significant digit of the indication.5Id for the indication of the values of both the standard and the unknown weights. A previous Type A evaluation of the repeatability of the measurement process. TW13 4UN Website: www.303log10 (1+x) ≈ For example: 0. with uncertainty limits of ± Id . for which limits were estimated to be ± 1 ppm of nominal value i. The least significant digit Id for the mass comparator represents 10 mg. and a weight of OIML Class F2.303log 10(1+0. A rectangular probability distribution has been assumed. K4.7 United Kingdom Accreditation Service.050 dB ± 0. the probability distribution for RF mismatch uncertainty is dependent on the combination of at least three mismatch uncertainties and can be treated as having a normal distribution. WX = WS + DS + δd + δ + W R + A b. Repeatability. UKAS Copyright exists in all UKAS publications PAGE 53 OF 82 . and are ± 30 mg. 3 K4 K4. This evaluation replicates the normal variation in positioning single weights on the comparator. The uncertainty evaluation has been carried out in accordance with UKAS requirements. where I C WS DS δd I δ C WR Ab = = = = = = Weight of the standard. Difference in comparator readings. No correction is made for air buoyancy.0227. NOTES 1 Combination of relatively small uncertainties expressed in dB is permissible since loge(1+x) ≈x when x is small and 2. The monitored drift limits for the standard mass have been set equal to the expanded uncertainty of its calibration.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K3. providing a coverage probability of approximately 95%. percentage. s(W R).5 g range permitted by the laboratory's procedures for the comparison was estimated from previous measurements to have limits of ± 3 mg.10 Reported result The measured value of the 30 dB attenuator at 10 GHz was 30. Feltham. Drift of standard since last calibration.ukas. e. Correction for air buoyancy. K4. For further details see paragraph E9. The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2.7 mg.049 dB. of 8. A rectangular probability distribution has been assumed. A rectangular probability distribution has been assumed.
com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. The uncertainty evaluation has been carried out in accordance with UKAS requirements. 2147 High Street.11 νor i ν eff Symbol Source of uncertainty Calibration of standard weight Drift since last calibration Digital rounding error. Middlesex.020 g = 10 000.005 g + 0.0 1.0 24.standard 1 2 3 + 0.02 g + 0. TW13 4UN Website: www.73 5. The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2. The calibrated value of the unknown is therefore 10 000.015 g + 0. providing a coverage probability of approximately 95%.0 3.015 g + 0.0 √ 3 √ 6 √ 3 √ 3 1.04 g + 0.0 5.0 17.025 g ± 0.020 g + 0. this is the value of n that is used to calculate the standard deviation of the mean: 8 . 3 K4.11 Reported result The measured value of the 10 kg weight was 10 000.01 g + 0. United Kingdom Accreditation Service.049 g. the mass of the standard is 10 000.025 g. UKAS Copyright exists in all UKAS publications PAGE 54 OF 82 .0 1. for which 10 readings were used (see paragraph B4).010 g Mean difference: + 0.9 Since three comparisons between standard and unknown were made (using 3 readings on the unknown weight).03 g + 0.0 Probability distribution Normal Rectangular Triangular Rectangular Rectangular Normal Normal Normal (k = 2) ui(W X) mg 15.32 4.ukas.0 ci 1.01 g + 0.10 Uncertainty budget Value ± mg 30.020 g From the calibration certificate.0 10.015 g + 0.0 mg.0 30.55 49.005 g.0 10.0 1. K4. u R s W X W = 5. Feltham.01 g Standard mean unknown .0 1.8 Three results were obtained for the unknown weight using the conventional technique of bracketing the reading with two readings for the standard. comparison Comparator nonlinearity Air buoyancy (1 ppm of nominal value) Repeatability of indication Combined standard uncertainty Expanded uncertainty Divisor 2.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K4.77 5.7 From Equations (5) and (6). NOTE The degrees of freedom shown in the uncertainty budget are derived from a previous evaluation of repeatability.03 g + 0. The results were as follows: No.0 1. Weight on pan standard unknown standard unknown standard unknown standard Comparator reading + 0.08 1.0 WS DS δd I δ C Ab WR u(WX) U ∞ ∞ ∞ ∞ ∞ 9 >500 >500 K4.025 g + 0.
s(WR ). The probability distribution is assumed to be rectangular. using the previously obtained repeatability standard deviation. The following uncertainty calculation is carried out for a near full loading of 200 g.4 K5. The probability distribution is assumed to be rectangular.e. n .2 mg. The span of the weighing machine has been adjusted using its internal weight before calibration. IX .005 mg.3 K5. Middlesex.1 mg and there is therefore a possible rounding error of ± 0. s (WR ).1. where I I WS DS δd0 I δd I Ab IR = = = = = = Weight of the standard. The rounding of the value of one digit of the indicated value. This can be thought of as an “internal digit” that is not presented to the user.e. The repeatability of the machine was established from a series of 10 readings (Type A evaluation).1 mg digit The calibration is carried out using weights of OIML Class E2.ukas. As the span of the machine was adjusted with its internal weight before calibration. The value of n that is used to calculate the standard deviation of the mean of the indication. i. Drift of standard since last calibration. and repeatability of the machine for repeated weighings near full load. TW13 4UN Website: www. The least significant digit on the range being calibrated corresponds to 0. The rounding of the value of one digit at the zero reading.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.1 mg.8 0 . NOTE This is a good example of how a detailed knowledge of the principles of operation of a measuring instrument may be required in order to identify and quantify the associated uncertainties. u s W X IR = 0. 2147 High Street. which gave a standard deviation. 1 United Kingdom Accreditation Service. the uncertainty limits were estimated to be ± 1 ppm of the nominal value. The probability distribution is assumed to be rectangular.05 mg. was 199. In this example.1 mg at a coverage probability of approximately 95% (k = 2).05 mg. eccentricity effects of the positioning of weights on the load receptor. K5. The resulting possible rounding error is therefore ± 0. it does so to a greater resolution than that provided by the digital readout.05 From Equations (5) and (6). or tared. UKAS Copyright exists in all UKAS publications PAGE 55 OF 82 . For this calibration point the weighing machine indication. No correction is made for drift. The degrees of freedom for this evaluation are 9.7 K5.2 The calibration certificate for the stainless steel 200 g standard mass gives an uncertainty of ± 0. No correction is made for air buoyancy.05 mg. Only one reading was taken to establish the weighing machine indication for each linearity and eccentricity point. Feltham. i.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K5 K5. Repeatability of the indication.1 Calibration of a weighing machine of 205 g capacity by 0. of 0. it is assumed that the “internal digit” size is onetenth of that provided on the external display. Correction for air buoyancy. is therefore one: K5.9999 g when the 200 g standard mass was applied. Checks will normally be carried out for linearity of response across the nominal capacity of the weighing machine. The machine indications are obtained from Unknown indication IX = W S + D S +δd0 + δd + Ab + IR . K5.6 K5. but the calibration interval is set so as to limit the drift to ± 0.5 No correction can be made for the rounding due to the resolution of the digital display of the machine. ± 0. It is often the case that when a weighing machine is zeroed.
003 0.0 WS DS δd0 I δd I Ab IR u(IX) U ∞ ∞ ∞ ∞ ∞ 9 >500 >500 K5. UKAS Copyright exists in all UKAS publications PAGE 56 OF 82 . 2147 High Street. providing a coverage probability of approximately 95%. m The change in value L D of the standard gauge block with time was estimated from previous calibrations to be zero with an uncertainty of ± 15 nm.0 1.9999 g ± 0.0 1. where L δ α T)] C LS LD δ L L α δ t δ α δ T DC δ C L V(x) Lr = = = = = = = = = = = = Certified length of the standard gauge block at 20ºC Drift with time of certified length of standard gauge block Measured difference in length Nominal length of gauge block Mean thermal expansion coefficient of the standard and unknown gauge blocks Difference in temperature between the standard and unknown gauge blocks Difference in thermal expansion coefficients of the standard and unknown gauge blocks Difference in mean temperature of gauge blocks and reference temperature of 20ºC when δ is L determined Discrimination and linearity of the comparator Difference in coefficient of compression of standard and unknown gauge blocks Variation in length with respect to the measuring faces of the unknown gauge block Repeatability of measurement K6. K6 K6.058 0.10 Reported result For an applied weight of 200 g the indication of the weighing machine was 199.1 0. TW13 4UN Website: www.30 mg.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.115 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K5.300 νor i ν eff Symbol Source of uncertainty Calibration of standard weight Drift since last calibration Digital rounding error (at zero) Digital rounding error (for indicated value) Air buoyancy (1 ppm of nominal value) Repeatability of indication Combined standard uncertainty Expanded uncertainty Divisor 2.9 Uncertainty budget Value ± mg 0. Feltham.005 0. with diminishing probability that the value approached the limits. From experimental evidence and prior experience the value of zero was considered the most likely.05 0. A triangular distribution was therefore assigned to this uncertainty contribution.3 The value of LS was obtained from the calibration certificate for the standard gauge block.0 1.2 K6. was determined from L X = LS + LD + δ – [L(α t + δ δ + DC + δ + L V(x) + Lr .1 0.0 √ 3 √ 3 √ 3 √ 3 1. The length of the unknown gauge block.0 ci 1.150 0.029 0.0 1.2 0.ukas.05 0. United Kingdom Accreditation Service. The uncertainty evaluation has been carried out in accordance with UKAS requirements. The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2. L X.1 Calibration of a Grade 2 gauge block of nominal length 10 mm The calibration was carried out using a comparator with reference to a grade K standard gauge block of similar material. The associated uncertainty was ± 0. Middlesex.05 0.0 1.03 μ (k = 2).05 Probability distribution Normal Rectangular Rectangular Rectangular Rectangular Normal Normal Normal (k = 2) ui(I X) mg 0.
6 nm. the resulting distribution was assumed to be triangular. Quadrature combination of these contributions gives an uncertainty due to surface irregularities of ± 30 nm. calculations based on the specifications for grade C gauge blocks indicted an uncertainty of ± 17 nm.ukas. DC . however.7 K6.6 K6.3. The difference δ between the mean temperature of the two gauge blocks and the reference temperature T of 20 ºC was measured and was assigned limits of ± 0. As the influence of δ αappears directly in both these uncertainty contributions they are considered to be correlated and. in accordance with paragraph D3.005 μ m. This Type A evaluation.2 ºC. was taken as zero with limits of ± 0. α m For L = 10 mm this corresponds to ± 20 nm/ºC. zero with limits of ± 0. United Kingdom Accreditation Service. Ts. This difference will have two influences: The difference in temperature. Feltham. the corresponding uncertainties have been added before being combined with the remaining contributions. 1 The measured length of the unknown gauge block was 9. Middlesex. 1 1 1 1 11. of similar magnitude. The variation in length of the unknown gauge block. Similarly. K6. based upon 11 measurements and using Equation (7). TW13 4UN Website: www.5 mm. As this is the combination of two rectangular distributions.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K6. α of .4 The coefficient of thermal expansion applicable to each gauge block was assumed to have a value.05 μ assessed from previous measurements. (ii) Effects due to surface irregularities such as scratches or indentations. because only one reading is made for the actual calibration. giving rise to a length uncertainty of ± 4 nm.5 μ m ºC with limits of ± 1 μ m ºC .x (a) (b) K6. giving rise to a length uncertainty of ± 1. δ between the standard and unknown gauge blocks was estimated to be t. such effects have a detection limit of approximately 25 nm when examined by experienced staff. The repeatability of the calibration process was established from previous measurements using gauge blocks of similar type and nominal length. UKAS Copyright exists in all UKAS publications PAGE 57 OF 82 . The calibration of the unknown gauge block was established from a single measurement. u R s L X L = 16 nm. is ± 2 μ m ºC with a triangular distribution.5 The error due to the discrimination and nonlinearity of the comparator. Combining these two rectangular distributions. with n = 1. L V(x). 2147 High Street. assuming this misalignment was within a circle of radius 0. δ . the difference in elastic compression δ m. the difference m m 1 1 in thermal expansion coefficient between the two blocks. yielded a standard deviation s(L R) of 16 nm.999 94 mm. C between the standard and unknown gauge blocks was estimated to be zero with limits of ± 0. This is included in the uncertainty budget as δ .8 16 From Equation (6). was considered to comprise two components: (i) Effect due to incorrect central alignment of the probe.08 ºC. as the conditions were the same as for the previous evaluation of repeatability the standard uncertainty due to repeatability can be obtained from this previous value of standard deviation.
THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K6.3 12.0 ci 1.0 1.0 5. The emfs generated by the thermocouples are measured using a digital microvoltmeter via a selector/reversing switch. 2147 High Street.1 Calibration of a Type N thermocouple at 1000ºC A Type N thermocouple is calibrated against two reference standard Type R thermocouples in a horizontal furnace at a temperature of 1000ºC.0 1. Feltham.0 √ 6 √ 3 √ 3 √ 6 √ 6 1.0 6.x T LV(x) Lr u(LX) U Source of uncertainty Calibration of the standard gauge block Drift since last calibration Comparator Difference in elastic compression Temperature effects Length variation of unknown gauge block Repeatability Combined standard uncertainty Expanded uncertainty Divisor 2. The temperature tx of the hot junction of the unknown thermocouple is given by K7.0 1. The uncertainty evaluation has been carried out in accordance with UKAS requirements.10 Reported result The measured length of the gauge block was 9.ukas.3 16. K7 K7.1 28. TW13 4UN Website: www.2 δ t t x ts iS δ iS1 δ iS 2 δ R 0 S V V V V CS0 δD tF δ t C t s iS s V iS1 s V iS 2 s V R s δ0 S δD δF V C δ C δ C δ t t t CS0 The voltage V x(t) across the thermocouple wires with the reference junction at 0 ºC during the calibration is t t 0 X t 0 X t Vx x t V tx iX V iX1 ViX 2 hTH V R V Cx CX 0 C X C X0 United Kingdom Accreditation Service.0 ∞ ∞ ∞ ∞ ∞ ∞ 10 >400 >400 K6.9 2. All the thermocouples have their reference junctions at 0ºC.9 2.9 νor i ν eff Symbol LS LD DC δ C δ s. UKAS Copyright exists in all UKAS publications PAGE 58 OF 82 .999 94 mm ± 0.9 Uncertainty budget Value ± nm 30 15 50 5.078 μ m. The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.0 1.6 30 16 Probability distribution Normal Triangular Rectangular Rectangular Triangular Triangular Normal Normal Normal (k = 2) ui(LX) nm 15. The unknown thermocouple is connected to the reference point using compensating cables.0 1.0 1. Middlesex. providing a coverage probability of approximately 95%.0 39.0 77.
δ iX2 V V δR V δ0S. The voltage sensitivity coefficients of the reference and unknown thermocouples have been obtained from reference tables as follows: Thermocouple reference unknown Sensitivity coefficient at temperatures of 0 ºC 1000 ºC CS = 0.8 K7. Sensitivity coefficients of the thermocouples for voltage at the measurement temperature of 1000 ºC. Temperature correction due to nonuniformity of the furnace.tX t δ LX V h TH K7. V V V Corrections were made to the microvoltmeter readings by using data from its calibration certificate.3ºC has been estimated from previous calibrations.6 K7.ukas. UKAS Copyright exists in all UKAS publications PAGE 59 OF 82 . Indication of the microvoltmeter.4 K7. . The measurement process consists of two parts . The temperature of the reference junction of each thermocouple is known to be 0 ºC within ± 0. TW13 4UN Website: www. δ0X t t C S. Voltage error due to the compensation leads.026 ºC/μ V K7. δ iX1 V V δ iS2 . of ± 0. Temperature at which the unknown thermocouple is to be calibrated. No correction is made for drift of the reference thermocouples since the last calibration but an uncertainty of ± 0.e. therefore only the calibration uncertainty of ± 2.0. Temperature corrections associated with the reference junctions. Error due to inhomogeneity of the unknown thermocouple.0 μ V. Drift of the reference thermometers since the last calibration.3 = = = = = = = = = = = = = = Temperature of the reference thermometer in terms of voltage with the cold junction at 0 ºC.407. A rectangular probability distribution has been assumed. the sensitivity coefficient associated with the uncertainty in the reference junction temperature is the ratio of those at 0 ºC and 1000 ºC. 2147 High Street. Voltage corrections due to the calibration of the microvoltmeter. The function is given in the calibration certificate.038 ºC/μ V CS0 = 0. rounding errors. V iX δ iS1 . The Type R reference thermocouples are supplied with calibration certificates that relate the temperature of their hot junctions with their cold junctions at 0 ºC to the voltage across their wires. For the 1000 ºC measurements.0 μ V (k = 2) is to be included in the uncertainty budget.077 ºC/μ V CX0 = 0.10 United Kingdom Accreditation Service. C X C S0. i.1 ºC.3 ºC with a coverage factor k = 2.189 ºC/μ V CX = 0.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. δ iS2 and δ iX2.7 The least significant digit of the microvoltmeter corresponds to a value of 1 μ This results in possible V. Voltage error due to contact effects of the reversing switch. Deviation of the temperature of the calibration point from the temperature of the furnace.5 μ for each indication. The reported result is the output emf of the test thermocouple at the temperature of the hot junction. Middlesex. CX 0 δD t δF t t Δ = t . The expanded uncertainty U is ± 0.9 K7. Rounding errors due to the resolution of the microvoltmeter.5 K7. K7. Drift and other influences were all considered negligible. Residual parasitic offset voltages due to the switch contacts were estimated to be zero within ± 2. Sensitivity coefficients of the thermocouples for voltage at the reference temperature of 0 ºC.determination of the temperature of the furnace and determination of the emf of the test thermocouple The evaluation of uncertainty has therefore been split into two parts to reflect this situation. Feltham.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 where tS(V) V iS .
11 The temperature gradients inside the furnace had been measured. This sequence reduces the effects of drift in the thermal source and parasitic thermocouple voltages.5 ºC. 1000.12 K7. The reported thermocouple output is 1000 V X 36248 µV = 36230 μ V.17 United Kingdom Accreditation Service.5 μ V 1000. By taking the mean they are combined into one observation of the temperature of the furnace at the location of the test thermocouple. one observation of the mean voltage of each thermocouple was deduced.3 ºC. These temperature values are highly correlated. TW13 4UN Website: www. Voltage differences between the leads and the thermocouple wires were estimated to be less than ± 5 μ V. UKAS Copyright exists in all UKAS publications PAGE 60 OF 82 . At 1000 ºC deviations from nonuniformity of temperature in the region of measurement are within ± 1 ºC The compensation leads had been tested in the range 0 ºC to 40 ºC.3 ºC.6 ºC Absolute mean values: Temperature of hot junctions: Mean temperature of furnace: K7. Four readings are thus obtained for all the thermocouples. Middlesex. The sequence of measurements is as follows: 1 2 3 4 5 First standard thermocouple Unknown thermocouple Second standard thermocouple Unknown thermocouple First standard thermocouple K7.16 K7.13 K7.4 ºC 1000. The error due to inhomogeneity of the unknown thermocouple was determined during the calibration by varying the immersion depth. The mean voltages of the reference thermocouples are converted to temperature observations by means of temperature/voltage relationships given in their calibration certificates. In a similar way one observation of the voltage of the test thermocouple is extracted. Corrections are not practical for this effect therefore the error was assumed to be zero within ± 0. The results were as follows: Thermocouple: Corrected voltages: First standard + 10500 μ V + 10503 μ V – 10503 μ V – 10504 μ V 10502.15 The thermocouple output emf is corrected for the difference between the nominal temperature of 1000 ºC and the measured temperature of 1000.ukas.0 μ V 1000. Feltham.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. If this is the case then the measurement must be repeated and/or the reason for the difference investigated.14 The polarity is then reversed and the sequence is repeated.6 ºC Unknown + 36245 μ V + 36248 μ V – 36248 μ V – 36251 μ V 36248 μ V Second standard + 10503 μ V + 10503 μ V – 10505 μ V – 10505 μ V 10504. K7. 2147 High Street.5 In this example it is assumed that the procedure requires that the difference between the two standards must not exceed 0. From the four readings on each thermocouple.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K7.
0 0.6 ∞ ∞ ∞ ∞ 9 ∞ ∞ >500 >500 United Kingdom Accreditation Service.0 √ 3 2.0 ºC Probability distribution Normal Rectangular Normal Rectangular Rectangular Normal Rectangular Rectangular Normal ui(T) ºC 0.0 √ 3 2.1 1.686 25.20 Uncertainty budget .1 ºC 0.7 2.emf of unknown thermocouple Value ± 0. A series of ten measurements had been undertaken at the same temperature of operation. Middlesex.0 1.0 1.0 μ V 2.1 ºC 0. K7.077 0.0 26.0 √ 3 √ 3 1.3 ºC 2.3 ºC 0.077 0.024 0.0 1.0 ∞ ∞ ∞ ∞ ∞ 9 ∞ ∞ >500 K7.6 μ V.077 1.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K7. this gave estimates of the standard deviations for the temperature of the furnace.641 νor i ν eff Symbol δ tS δD t δ iS1 V δR V δ t0S tS δ iS2 V δ tF uc(T) Source of uncertainty Calibration of standard thermocouples Drift in standard thermocouples Voltmeter calibration Switch contacts Determination of reference point Repeatability Voltmeter resolution Furnace nonuniformity Combined standard uncertainty Divisor 2.0 μ V 2.0 1.173 0.0 √ 3 √ 3 ci 1.9 νor i ν eff >500 Symbol ΔX t δ LX V δ iX1 V δR V δ t0X u(ViX ) δ iS2 V hTH uc(V) U Source of uncertainty Furnace temperature Compensating leads Voltmeter calibration Switch contacts Determination of reference point Repeatability Voltmeter resolution Inhomogeneity of thermocouple Combined standard uncertainty Expanded uncertainty Divisor 1.6 μ V.0 μ V 2.10 0.6 0.1 ºC 1.0 μ V 0.0 √ 3 √ 3 ci 38. n 1 The value of n = 1 is used to calculate the standard uncertainty because in the normal procedure only one sequence of measurements is made at each temperature.6 μ V 0.5 μ V 1.089 0.1 tS From Equation (6): u sp t S tS = 0.29 6.0 μ V 0.641 ºC 5.0 38.155 1. n 1 s p iX 1.6 V and u iX sp V iX V = 1. of 1.19 Uncertainty budget .577 0.0 0.temperature of the furnace Value ± 0.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Feltham.6 1. UKAS Copyright exists in all UKAS publications PAGE 61 OF 82 .9 51. sp (tS) of 0.5 μ V 0.3 ºC Probability distribution Normal Rectangular Normal Rectangular Rectangular Normal Rectangular Rectangular Normal Normal (k = 2) ui (V) μ V 24.407 1.10 ºC.0 √ 3 √ 3 1.0 1. 2147 High Street.150 0. TW13 4UN Website: www. sp (V iX ).022 0.18 In order to determine the random uncertainty associated with these measurements a Type A evaluation had been carried out on a previous occasion. The resulting standard uncertainties were as follows: s p 0 .10 ºC and the voltage of the thermocouple to be calibrated.48 1.ukas.89 1. Using Equation (7).077 0.
2147 High Street. The indication of the unknown DPI is obtained from: K8.1 United Kingdom Accreditation Service. K8 Calibration of a Digital Pressure Indicator (DPI) at a nominal pressure of 2 MPa using a reference hydraulic dead weight tester The pressure was generated using a dead weight tester (DWT) the performance characteristics of which had previously been determined.2 ppm.5 K8. ApD. 2 2 K8.5. The indication was approached with increasing pressure to account for the existence of possible hysteresis in the DPI.0 and with its cold junction at a C temperature of 0 ºC.6516 mm ± 0. TW13 4UN Website: www. based on results from previous calibrations.2 = = = = = = = = = = = = = = = = Indication of unknown DPI Mass of the component parts of the load. an emf of 36 230 μ ± 52 μ V V. The drift limit in the effective area of the DWT.0026 mm . Density of hydraulic fluid Surface tension coefficient of hydraulic fluid Circumference of reference piston Effective area at zero pressure of reference piston Distortion coefficient of reference piston (pressure dependant term) Temperature coefficient of piston and cylinder The rounding of the value of one changing digit of the indication Repeatability of indication Error due to the piston not being perfectly vertical.3 The calibration certificate for the reference DWT gives the distortion coefficient of the reference piston 6 6 as ap = 6. has been set to ± 30 ppm.0 x 10 /MPa ± 0.5 x 10 /MPa. including the piston Density of ambient air Density of the mass. This results in a relative uncertainty in A0 of 32.21 Reported result The type N thermocouple shows. providing a coverage probability of approximately 95%. Feltham. at the temperature of 1000. The calibration certificate for the reference DWT gives the piston area and its uncertainty as: A0 = 80.000 010 kg. UKAS Copyright exists in all UKAS publications PAGE 62 OF 82 . The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2. and can be significantly different for each load component.ukas.4 K8. Middlesex.567 227 kg ± 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K7.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Mass uncertainties The mass of the piston is shown on the calibration certificate as 0. The value of the local acceleration due to gravity Height different between reference level of standard and reference level of generated pressure Buoyancy volume of the reference piston – from calibration certificate. K8. m.1 IX 1 mL g a BV f a g C m f d IR e V a gh I A 0 1 a p t 20 1 where IX mL ρ a ρ m g h BV pf σ C A0 ap λ δd I IR eV K8. The uncertainty evaluation has been carried out in accordance with UKAS requirements.
60 ppm. as all masses have an assumed density of 3 7800 kgm . The expanded uncertainty associated with this estimate. fh .6 The uncertainty in the temperature of the piston.402 137 . No correction is made therefore a limit value of ± 3 mm has been assigned for the uncertainty associated with fluid head effects. A correction had been made for the value of the local acceleration due to gravity. from buoyancy volume.4 K8. Assuming that the density of the oil used 3 2 is 917 kg/m and the local value of g is 9. It is estimated as +2 mm and the uncertainty in this estimate is ± 1 mm.e.224 784 0 7 8 Similarly. i. the uncertainty.3 The drift of the piston mass. is estimated to be ± 13 ppm. In relative terms this corresponds to an uncertainty of ± 13.8 K8.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. No correction has been made for air buoyancy therefore. surface tension and fluid density effects has been estimated as ± 2 ppm based on a relative uncertainty in each of ± 10%.000 015 kg.003 = 27. coming from the analysis of the temperature indictor used has been set as ± 0. u t. 10 mg.0 Pa. The repeatability of the calibration process had been established from previous measurements using DPIs of similar type and nominal range. This value had been estimated from knowledge of the measurement location and took the Bouguer anomalies into account.81 x 0.9 K8. Ug. in relative terms: Relative uncertainty in mass u(m) = Uncertainty of Piston (mg) A B C ppm Mass of Piston A (kg) B C = 10 10 50 50 = 7.402 137 kg ± 0. ab . arises from the height different between the reference level of the reference DWT and the generated pressure datum point.30 ppm 0. based upon 10 measurements and using Equation (7). K8.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K8.5.5.000 050 kg C = 8. UKAS Copyright exists in all UKAS publications PAGE 63 OF 82 . yielded a relative standard deviation s(PR ) of 16 ppm.567 227 .11 United Kingdom Accreditation Service. fe. The uncertainty contribution. Feltham. The uncertainty of the mass set is shown on the calibration certificate for the three weights “A”. has been set to 0.000 050 kg K8.255 242 kg ± 0.224 784 kg ± 0. To obtain the combined mass uncertainty. This will affect the pressure generated in proportion to the temperature coefficient of the piston and cylinder combination. The uncertainty relating to fluid head effects.81 ms .5. In this case a steel piston and cylinder has a temperature coefficient of 23 ppm/ºC. TW13 4UN Website: www. This Type A evaluation.10 K8.5 ppm at 2 MPa. was ± 3 ppm (k = 2). K8.7 K8. This figure was obtained from the calibration certificate for the DWT. “B” and “C” used to generate 2 MPa as: A = 0. based on previous calibrations. then the uncertainty associated with the fluid head effect is 917 x 9.5. 50 mg and 50 mg respectively.255 242 .ukas. for drift this gives mD = 7.5ºC.000 010 kg B = 7.2 K8. 2147 High Street. Middlesex.5 The limits to the drift of the mass set have been set to be equal to the expanded uncertainty of its calibration. u(m).
4 ppm of the generated pressure.0 δd I IR u(IX) U United Kingdom Accreditation Service. K8. In this example.50 7.0 ppm 13.0 2. As the conditions were the same as for the previous evaluation of repeatability the standard uncertainty due to repeatability can be obtained from this previous value of standard deviation.0 1.0 86.15 Uncertainty budget νor i ν eff ∞ ∞ ∞ ∞ ∞ ∞ ∞ ∞ ∞ ∞ ∞ ∞ 9 >350 >350 Symbol Source of uncertainty Value ± 32.2 ppm 0. because only one reading is made for the actual s R P calibration.0 1.ukas.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K8.999 996 6. after levelling. K8.6 ppm 0. The probability distribution is assumed to be rectangular.64 7. 2147 High Street. As this uncertainty is small compared with others in this particular calibration.15 1.51 1.79 1.0 1.000 12 MPa for an applied pressure of 2 MPa.0 1.14 K8.96 28. NOTE This effect always acts in one direction.12 An uncertainty arises due to the fact that the piston may not be perfectly vertical.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.0 1. then all of the force would act on the area. The I least significant digit on the range being calibrated of this particular DPI changes in steps of 200 Pa and there is therefore a possible rounding error of ± 100 Pa or. Any departure from vertical will reduce the force .0 A0 ap ApD (u)m mD ut ab Ug fh fe uL Calibration of DWT (area) Calibration of DWT (distortion) Drift in area Calibration of total load Drift of total load Temperature of the piston Air buoyancy Local gravity determination Fluid head effects Other fluid effects Levelling effects Digital rounding error Repeatability of indication Combined standard uncertainty Expanded uncertainty 2.500 17. the piston is vertical to within 0.by the cosine of the angle. UKAS Copyright exists in all UKAS publications PAGE 64 OF 82 .15°.0 43. it is convenient to treat it as bilateral.3 ppm 7. TW13 4UN Website: www.0 1. Feltham. ± 50 ppm. with n = 1.5 ppm/MPa 30 ppm 7.25 0. The maximum error is therefore 3.15° is 0. δd. 1 No correction can be made for the rounding.13 The calibration of the unknown DPI was established from a single measurement.0 1.and therefore the pressure .0 ppm 3. due to the resolution of the digital display of the DPI. it is assumed that.0 2.32 3. from which the DPI indication was 2.4 ppm 50 ppm 16 ppm Probability distribution Normal Normal Rectangular Normal Rectangular Rectangular Rectangular Normal Rectangular Rectangular Rectangular Rectangular Normal Normal Normal (k = 2) Divisor ci ui(PX) ppm 16.65 4.e.0 1. Middlesex. i.5 ºC 13 ppm 3. The effect in terms of generated pressure is proportional to the cosine of the angle from true vertical.0 √ 3 2 √ 3 √ 3 √ 3 2 √ 3 √ 3 √ 3 √ 3 1 1. the generated pressure will always be smaller than that obtained if the piston were truly vertical.0 23 1. Therefore IR = = 16 ppm.0 1.9 16. If it were. The cosine of 0.5 ppm 2. in relative terms.38 6.
5. However. If it is required that the uncertainty is reported in absolute units.0315 N/m This results in an applied pressure of 2.17 Reported result The pressure was applied in an increasing direction until it reached a final value of 2.000 17 MPa.3) and the following quantities: Temperature of Piston: Buoyancy volume of piston: Air density: Local gravity: Oil surface tension coefficient: 21 ºC 3 3.ukas. This has been verified by Monte Carlo Simulation techniques.000 84 MPa ± 86 ppm. therefore the digital rounding error is 1 digit. 2147 High Street. that of the mass set (K8. the expanded uncertainty in absolute terms is ± 0.811812 ms 0. Middlesex. The reported expanded uncertainty is based on a standard uncertainty multiplied by a coverage factor k = 2. UKAS Copyright exists in all UKAS publications PAGE 65 OF 82 . This uncertainty budget has been constructed in relative terms (ppm). even though they are of smaller magnitude. K8. the combined standard uncertainty is still Gaussian. δd. this changes in steps of 2 digits for this particular DPI. The resolution uncertainty is based on the least significant digit of the DPI.000 806 MPa. it can be calculated from the reported value and the relative uncertainty.107 m 3 1. Feltham.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 K8. The uncertainty evaluation has been carried out in accordance with UKAS requirements. is larger than any other contribution and is assigned a rectangular I distribution.1). The indication of the digital pressure indicator was 2. NOTES 1 In this example. In this case. providing a coverage probability of approximately 95%.16 The generated pressure was calculated from the mass of the piston (K8.5. due to the presence of the other uncertainties. TW13 4UN Website: www.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. the uncertainty due to resolution.2 kg/m 2 9. as most of the errors that arise are proportional to the generated pressure and it is the convention in this particular field of measurement to express uncertainties in this manner. Nevertheless.000 806 MPa. 2 3 United Kingdom Accreditation Service.
2 L2. of course. Traditionally this has been expressed in the form ± (UREL + UABS) This linear addition of quantities is not in accordance with the principles embodied in the GUM. when reporting values higher up the scale it may be desirable to express the uncertainty in relative terms. Principles When measurements are made over a range of values and the corresponding sources of uncertainty are examined it may be found that some are absolute in nature (i. in effect.4 L2. UKAS Copyright exists in all UKAS publications PAGE 66 OF 82 . Middlesex.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX L EXPRESSION OF UNCERTAINTY FOR A RANGE OF VALUES L1 L1.5 United Kingdom Accreditation Service. they arise in a manner that makes them proportional to the value of the measurand). they arise in a manner that is independent of the value of the measurand) and some are relative in nature (i. The GUM deals with expression of uncertainties for the reporting of a single value of a measurand. the user may also require to express these uncertainties in both absolute and relative terms. to calculate a value for the expanded uncertainty for each reported value over the range.13.1 L2. If the instrument being calibrated is subsequently to be used in a situation where a further analysis of uncertainty is required. A suggested statement is given in the example in L3. The two values should normally be reported separately with an appropriate statement describing how they should be combined. unless there happens to be a high degree of correlation between the absolute and relative terms .which is not usually the case. or more than one parameter derived from the same set of data. Additionally.3 L2.1 Introduction On occasions it is convenient to provide a statement of uncertainty that describes a range of values rather than a single result. L1.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. It is possible. These evaluations are carried out in the manner already described in this publication.e. This can give problems when reporting values near zero as a relative term may not be appropriate and an absolute term has to be used. a separate uncertainty evaluation is carried out for each. The results of these evaluations are then expressed as separate absolute and relative terms.2 L1. TW13 4UN Website: www. explains how it can be dealt with using the principles of this code of practice and provides an illustration of the process using a worked example. This Appendix therefore describes the situations when this can occur.ukas. Feltham. Reporting uncertainties in both absolute and relative terms therefore provides more information to the user than if a series of single values are quoted. 2147 High Street. as this is often how instrument specifications are expressed. it would not be possible to extract the absolute and relative parts from these single values. However if the user has only been provided with a single value of uncertainty for each reported value. The process of calculating a value of expanded uncertainty describing a range of values is identical to that for single values except that the absolute and relative terms are identified as such and.e. it is often more representative of the way instrument specifications are expressed. Conversely.3 L2 L2. In practice many measuring instruments are calibrated at several points on a range and the use of an expression describing the uncertainty at any of these points can be desirable.
No corrections were made for known errors of the calibrator as these were identified as being small relative to other sources of uncertainty. Drift in voltage of multifunction calibrator since last calibration. Effects on voltmeter reading due to imperfect commonmode rejection characteristics of the measurement system.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. V CM. A value of ± 1 μ was assigned. TW13 4UN Website: www. was assigned. calibrator and multimeter. The effects of thermoelectric voltages.5 μ in absolute units (k = 2).8 L3. Linearity and zero offset of multifunction calibrator.8 ppm as a relative uncertainty but there was an additional ± 0. where V V CAL VD V UE V TC V LIN VT V CM δ RES V IR = = = = = = = = = Calibrated voltage setting of multifunction calibrator. The calibrations were carried out in both polarities at 0. The probability distribution was assumed to be rectangular.2 L3.9 United Kingdom Accreditation Service.5 V and 1. IDVM . based on previous experiments with V the leads. A rectangular distribution was assumed. This had a value of ± 2. Feltham.6 L3.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 L3 L3.3 L3. 2147 High Street. Middlesex. had also been the subject of a previous evaluation and a value of ± 1 μ with a rectangular distribution. using internal quality control checks and ensuring that it was used within the temperature range and other conditions as specified by the manufacturer. Thermoelectric voltages generated at junctions of connecting leads. L3. V The manufacturer's 1year performance specification for the calibrator included the following effects: V D . Repeatability of indication L3. Temperature coefficient of multifunction calibrator. Only one measurement was carried out at each point and therefore reliance was placed on a previous evaluation of repeatability using similar multimeters. Uncorrected errors of multifunction calibrator. the commonmode voltage is unrelated to the measured voltage.1 Example of uncertainty evaluation for a range of values In this example a 6½digit electronic multimeter is calibrated on its 1 V dc range using a multifunction calibrator.9 V. V LIN This contribution was assumed to be absolute in nature. therefore the absolute 6 term is ± 2 V x 10 = ± 2 μ The performance of the calibrator had been verified by examining its V. can be described as follows: IDVM = V CAL + VD + V UE + VTC + V LIN + VT + V CM + δ RES + IR . UKAS Copyright exists in all UKAS publications PAGE 67 OF 82 . This contribution is absolute in nature. Rounding errors due to the resolution of multimeter being calibrated. VT. The indication of the multimeter under test. as V. Thermoelectric voltages are independent of the voltage setting are therefore an absolute uncertainty contribution. The uncorrected errors are assumed to be zero with an uncertainty obtained by analysis of information obtained from the calibration certificate for the calibrator.4 L3.5 The calibration uncertainty was obtained from the certificate for the multifunction calibrator. calibration data and history.ukas. V TC These contributions were assumed to be relative in nature. V UE . On this particular multifunction calibrator the fullscale value is twice the range value. L3. for the particular connecting leads used had been evaluated on a previous occasion.7 The specification for the calibrator on the 1 V dc range was ± 8 ppm of reading ± 1 ppm of fullscale.1 V increments from zero to 1 V and additionally at 1. Effects due to commonmode signals.
0 value (absolute) ±μ V 0.9 V.0 1.5 5.ukas. After the results the following statements regarding uncertainty can be given: The expanded uncertainty for the above measurements is stated in two parts: Relative uncertainty: ± 11 ppm Absolute uncertainty: ± 2.58 0. this term is absolute in nature.15 0.5 Probability distribution Normal Rectangular Rectangular Rectangular Rectangular Normal Normal Normal (k = 2) ui(V) (relative) ppm 1.42 10.5 2.0 1.0 ci 1 1 1 1 1 1 L3. The least significant digit on the range being calibrated corresponds to 1 μ and there is therefore a V possible rounding error δ RES of ± 0. 2147 High Street. From Equation (7). n 1 L3.58 0.5 2. Repeatability at the zero scale point was found not to be significant compared with other absolute contributions. Middlesex.62 ui(V) (absolute) μ V 0.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 L3.0 √ 3 √ 3 √ 3 √ 3 1.10 No correction is made for the rounding due to the resolution VRES of the digital display of the multimeter. Feltham.12 Uncertainty budget value (relative) ± ppm 2. The uncertainty evaluation has been carried out in accordance with UKAS requirements.5 μ The probability distribution is assumed to be rectangular and V V.8 1. if required.13 It is assumed that the results of this calibration will be presented in tabular form. TW13 4UN Website: www.40 4.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications PAGE 68 OF 82 . combine the uncertainties shown by quadrature summation in either relative or absolute terms as appropriate. this was found to give a standard deviation s(V i) of 2.92 νor i ν eff ∞ ∞ ∞ ∞ ∞ 9 >100 >100 Symbol VCAL VSPEC VT VCM δ RES V IR uc (V) U Source of uncertainty Calibration uncertainty Specification of calibrator Thermoelectric voltages Common mode effects Voltmeter resolution Repeatability Combined standard uncertainty Expanded uncertainty Divisor 2.5 ppm. Ten measurements were carried out at zero voltage.5 ppm.9 μ V The reported twopart expanded uncertainty is in each case based on a standard uncertainty multiplied by a coverage factor k = 2.11 A previous evaluation had been carried out on the repeatability of the system using a similar voltmeter.25 1. For each stated result the user may. United Kingdom Accreditation Service. providing a coverage probability of approximately 95%. Then.0 0. s 2 . 1 V and 1.46 2. L3.5 V IR sp V i i = 2. from Equation (6).29 2.8 8.
2 M2. it can be the case that one. Middlesex. M1. For calibration activities. and specification compliance is less relevant. or both.1 M2. For measurement standards it is more likely that the measured value and expanded uncertainty will be of more interest to the user. thereby limiting the amount by which the tails of the distribution extend away from the reported value. If the result y and the entirety of the distribution lie outside the specified limits then noncompliance with the specification has been demonstrated. Nevertheless. or both.ukas. A “perfect” normal distribution has tails that extend to infinity in each direction. Conversely.3 M2 M2. If the result y and the entirety of the distribution lie within the specified limits then it is clear that compliance with the specification has been demonstrated. Feltham. Theory It has been shown elsewhere in this document that an uncertainty associated with a measurement result will usually take the form of a normal or Gaussian distribution centred around the reported value y of the measurand Y. of the tails of the distribution overlapping the limits.4 M2. of the tails of the distribution significantly overlap one. the implication of this is that compliance – or noncompliance – can only be stated in conjunction with an associated confidence level. implying that there will always be some doubt about whether or not compliance has or has not been demonstrated. this will often be the case for general purpose test and measurement equipment. This means that consideration has to be given to the area of the distribution that is contained within the limits when assessing compliance with the specification.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications PAGE 69 OF 82 . Furthermore.3 M2. In practice. 2147 High Street. or both. This concept is illustrated in Figure 7 overleaf. of the specification limits. an uncertainty budget is always based on a finite number of contributions.6 United Kingdom Accreditation Service.2 Introduction In many situations it will be necessary to make a statement in calibration certificates or test reports about whether or not the reported result complies with a given specification.1 M1.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX M ASSESSMENT OF COMPLIANCE WITH SPECIFICATION M1 M1. This is because there will always be a possibility of one.5 M2. For testing activities it is very likely that the result will have to be compared with specified limits in order to arrive at a conclusion relating to conformance or fitness for purpose. TW13 4UN Website: www.
11 M2. with the specification. using more accurate equipment or by taking the mean of a larger amount of readings.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. with a coverage probability of approximately 95%. This will always yield a “safe” decision. Similar reasoning applies regarding noncompliance with the specification if the result were to be outside the limits. it is generally accepted practice that statements regarding compliance will relate to the same level of confidence. NOTE It is common practice for a compliance decision to be made by simply comparing the expanded uncertainty. the uncertainty is small compared with the specification.ukas. Middlesex.8 As an expanded uncertainty is normally expressed for a coverage probability of approximately 95%. with a specification cannot be demonstrated at 95% confidence. This is because only one tail of the distribution will usually be the subject of comparison with the limit. It will yield a confidence level of 97. as described above.5% or greater.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 Figure 7 M2. M2.64. be taken with the agreement and understanding of the customer. then compliance has been demonstrated at a confidence level of 95%.10 M2. This procedure may also be used in cases where a customer has requested a compliance statement for a confidence level other than 95%. as should be the case. possibly by applying corrections to the result. it can be seen that the majority of the distribution lies within the specification limits. of course. or noncompliance.9 M2. TW13 4UN Website: www. should therefore be used for the purpose of comparison with a specification limit. One solution is to reduce the uncertainty. If this is not practical or desirable. M2. A new coverage factor. If. and 5% of it is outside them. 95% of the area of the distribution is within the specification limits. the probability contained within the other tail will already be within the specification limits. UKAS Copyright exists in all UKAS publications PAGE 70 OF 82 . however the method described above is preferred as it takes into account the actual amount of probability that breaches the specification limit. then it may be possible to evaluate compliance. the value of ks required to achieve at least 95% confidence is 1. Assuming a normal distribution. ks. 2147 High Street.12 It may be the case that compliance. If. Feltham. Such an approach should. or noncompliance. at a different level of confidence. for example. If this is the case for a given situation.13 United Kingdom Accreditation Service. then comparison of the expanded uncertainty using a coverage factor k = 2 with the specification limit is unduly pessimistic.7 Lower specification limit Result Upper specification limit In this example. but a significant proportion lies above the upper limit. M2.
Alternatively.28. if the result coincided exactly with one of the limits.78 2.88 1.16 LS y u c y 3.45 95 94 93 92 M2.48 1. M2. can be obtained from the following table: Probability of compliance or noncompliance M2.y M2.08 1.95 0.52 99.71 0. The specification is ± 1.32 2. Furthermore. The next lower value in the table is 1.68 0. United Kingdom Accreditation Service.29 2.74 0. as the result approaches either limit there will come a point at which no reasonable decision can be made regarding compliance or noncompliance with the specification.15 The probability of compliance when the result lies within the specification. At what confidence level can compliance with the specification be made? L S y u c y = 1 . higherorder interpolation should be used otherwise.23 1.00 .15 units.00 unit. Linear interpolation will suffice for LS y u c y <2. or that of noncompliance when it lies outside the specification.77 0.81 0.92 0.58 0.75 1.17 Normally 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 72 71 70 LS y u c y will not be an integer and it will be necessary to interpolate between the values given in the table.15 = 1.28 1. therefore it has been demonstrated that the specification is met for at least 90% confidence.84 0.13 1.69 1. 2147 High Street.04 1.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 M2.80 units with a combined standard uncertainty u c(y) of 0.33. the table above is limited to a confidence probability of 70% and above.18 It should be noted that this procedure is only valid when the uncertainty breaches one of the specification limits and for this reason the uncertainty should be sufficiently small that an insignificant portion of the distribution approaches the other limit.00 0. Middlesex. M2. LS . regardless of the magnitude of the uncertainty.56 1. For this reason.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.55 0. TW13 4UN Website: www.14 Two pieces of information are needed to deduce the confidence level at which compliance can be stated: The combined standard uncertainty u c(y) The difference between the specification limit and the result.19 M2.73 99 98 97 96 95.88 Probability of compliance or noncompliance LS y u c y 0.ukas.80 0 0 . and by general convention.34 1.41 Probability of compliance or noncompliance LS y u c y 1.9 99. there would always be 50% confidence in the decision. UKAS Copyright exists in all UKAS publications PAGE 71 OF 82 . In the extreme. the next lower value may be used. Feltham.05 1.64 0.17 1.20 Example A measurement yields a result y of 0.61 0.64 1.
If both the uncertainty U and the specification L are stated at the same coverage probability. Feltham. 45 k99 2 . 2.45 10 = 7. The specification at the 95. taking the measurement uncertainty into account. An example of how this may be reported is given below: The measured result is below (above) the specification limit by a margin less than the measurement uncertainty.1 There are situations where the specification is characterised by a normal distribution. Middlesex. U S M3. U. Specification limits – further considerations In Section M2 the specification limits have been treated as absolute.22 M2. can be k obtained from L95 .THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 M2. 2147 High Street.2 L2 2 and is failed when y > U S L2 2 . it is not permitted in some areas of legal metrology. then a normal distribution can be assumed.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. This may not always be the case.47.58 L2 . The expanded uncertainty. TW13 4UN Website: www.45 = 2 and k99 = 2. when a specification is quoted for a given coverage probability.000 000 V dc. UKAS Copyright exists in all UKAS publications PAGE 72 OF 82 . Is the specification met? The comparison with specification has to be carried out with both the specification and the uncertainty at the same coverage probability.58. Some manufacturers state confidence levels for their specifications. k95.0 From Paragraph 3. 45 L99 95 . M3 M3. then compliance with the specification has been demonstrated. As 7 < 7. is ± 3 ppm (approximately 95% confidence. analogous to a rectangular probability distribution. k = 2) and the reading is + 7 ppm from the nominal value. for example. It is stated in the GUM that. 45 95. therefore L95 . United Kingdom Accreditation Service.75 2 3 2 = 7. It should also be noted that this “shared risk” approach may be superseded by legal requirements.23 In cases such as those described above it is essential that the client is made aware of the situation because the enduser is taking some of the risk that the item may not meet the specification. However the result indicates that compliance (noncompliance) is more probable than noncompliance (compliance) with the specification limit. L95. it is therefore not possible to state compliance (noncompliance) based on the stated coverage probability. The manufacturer’s specification for this reading is stated as 10 ppm at 99% confidence. then the specification is met at that coverage probability when y < where y = reported result U = expanded uncertainty LS = specification limit A worked example is shown below. but only a statement of the likelihood of compliance is required. M2.15 ppm.45% coverage probability.21 It may be the case that a result cannot be demonstrated to comply with a specification for a given confidence level.45 .15.75 ppm.3 Example A digital multimeter is calibrated with an applied voltage of 10 .ukas. 45 = U2 95 7 . M3.
TW13 4UN Website: www. M4.2 M4. United Kingdom Accreditation Service. the specification and any relevant clauses within it should be unambiguously identified in the calibration certificate or test report.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. However care must be taken to ensure that there is no implication that parameters that have not been measured also comply with a specification. due allowance having been made for the uncertainty of the measurements.ukas. There may be cases where the uncertainty attainable for a given test or calibration is larger than the specification for the item under consideration. This statement can be modified as necessary where noncompliance with a specification is to be reported. UKAS Copyright exists in all UKAS publications PAGE 73 OF 82 . For this reason a broad statement such as "the equipment complies with its specification" should not be made. A suggested statement of compliance is as follows: The equipment complies with the stated specification at the measured points for the stated confidence level.1 Reporting compliance with specification If compliance or noncompliance with a specification is clearly demonstrated for a given confidence level then a statement to this effect can be made in calibration certificates and test reports.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 M4 M4. Middlesex. Feltham.3 When making compliance or noncompliance statements. Such situations should be subject to the contract arrangements between the laboratory and its customer. 2147 High Street. A statement regarding compliance should not be given under these circumstances but a note should be included indicating those uncertainties associated with the measured values that are greater than the required specifications.
Testing laboratories should therefore have a defined policy covering the evaluation and reporting of the uncertainties associated with the tests performed. 2147 High Street.2 A measurement begins with an appropriate specification of the measurand. the strength of a material. which may differ from one field to another.ukas. the quantity of microorganisms in a food sample. to which much of this document is addressed. or requirements of specifications. Some tests are qualitative in nature.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX N UNCERTAINTIES FOR TEST RESULTS N1 N1. the susceptibility of an appliance to electric or magnetic fields.1 N1. treatment and reporting of the uncertainty. However laboratories should be able to satisfy requests from clients.4 N1. It is recognised that the present state of development and application of uncertainties in testing activities is not as comprehensive as in the calibration fields. for example: the electrical breakdown characteristics of an insulating material. the generic method of measurement and the specific detailed measurement procedure. TW13 4UN Website: www. the level of emissions of electromagnetic radiation from an appliance.2 Introduction ISO/IEC 17025:2005 requires that “testing laboratories shall have and apply procedures for estimating uncertainty of measurement”.e. they do not yield a numeric result. Knowledge of the influence quantities involved for a given procedure is important so that the sources of uncertainty can be identified. When applied to testing. Nevertheless. the general term measurand may cover many different quantities. the quantity of asbestos particles in a sample of air.1 United Kingdom Accreditation Service. The laboratory should use documented procedures for the evaluation.. Objectives The objective of a measurement is to determine the value of the measurand.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. It is therefore accepted that the implementation of ISO/IEC 17025:2005 criteria on this subject will take place at an appropriate pace. The methodology for estimation of uncertainty in testing is no different from that in calibration and therefore the procedures described in this document apply equally to testing results. Therefore there can be no meaning in reporting uncertainties directly associated with the test result. i. there will be uncertainties associated with the underlying test conditions and these should be subject to the same type of evaluation as is required for quantitative test results. to provide statements of uncertainty. UKAS Copyright exists in all UKAS publications PAGE 74 OF 82 . ie the specific quantity subject to measurement.3 N1.5 N2 N2. N2. Feltham. the concentration of an analyte. N1. Middlesex.
such as microbiological testing. Previous measurement data . Sampling . (f) (g) Values assigned to measuring equipment and reference materials. However the following general points will apply to many areas of testing: Incomplete definition of the test . unrecognised systematic effects may exist that cannot be taken into account but contribute to error. . e. 2147 High Street.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. Middlesex.the sample may not be fully representative. humidity and air pressure.g. Imperfect realisation of the test procedure.Reaction time. As these will depend on the nature of the tests involved.3 (a) (b) United Kingdom Accreditation Service.the requirement may not be clearly described. In some disciplines. e. (a) (b) (c) (d) (e) (h) (i) (j) N3. the temperature of a test may be given as 'room temperature'.Instrument resolution or discrimination threshold. TW13 4UN Website: www. This is one reason that participation in interlaboratory comparisons. history graphs can be constructed and can yield useful information about changes with time. variability in the performance of the person carrying out the test and variability in the homogeneity of the sample itself. it can be very difficult to obtain a representative sample. even when the test conditions are clearly defined it may not be possible to produce the theoretical conditions in practice due to unavoidable imperfections in the materials or systems used.g. Feltham. Information on some of the sources of these errors can be obtained from: Data in calibration certificates . e. . . Variations in repeated observations made under similar but not identical conditions . when using a stopwatch.ukas. Inadequate knowledge of the effects of environmental conditions on the measurement process. UKAS Copyright exists in all UKAS publications PAGE 75 OF 82 .this enables corrections to be made and uncertainties to be assigned. temperature.g. it is not possible to give detailed guidance here. Values of constants and other parameters used in data evaluation.Judgement of colour. Changes in the characteristics or performance of measuring equipment or reference materials since the last calibration. shortterm fluctuations in the local environment.2 N3. for example: .such random effects may be caused by. These sources are not necessarily independent and. electrical noise in measuring instruments. in addition.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 N3 N3.Reading of scales on analogue indicating instruments. Approximations and assumptions incorporated in the measurement method and procedure. Personal bias and human factors. or errors in graduation of a scale. participation in proficiency testing schemes and internal crosschecking of results by different means are encouraged. or imperfect measurement of environmental conditions.for example. for example.1 Sources of uncertainty There are many possible sources of uncertainty.
In such cases. However it will sometimes be the case that the procedure requires standard reference materials to be subject to the same process. matrix effects may also have to be considered. (d) (e) (f) N3. Take the independent components and the values of any derived net components and. account for whether they act together or cancel each other) and derive a net value. if this is the case then the evaluation is designated Type A. 2147 High Street. such instruments have to be shown to comply with the relevant specifications. normally by calibration. This gives the combined standard uncertainty (Equation (1)). Feltham.3. Express each uncertainty value as the equivalent of a standard deviation (see paragraph 3. Assign values to these using information such as described in N3. UKAS Copyright exists in all UKAS publications PAGE 76 OF 82 . Middlesex.28 to 3. In some fields of testing it may be the case that the contribution of measuring instruments to the overall uncertainty can be demonstrated to be insignificant when compared with the repeatability of the process.35). for example there may be various stages of weighing. Accepted values of constants associated with materials and quantities. or in the case of Type A evaluations.. Nevertheless. calculate the standard deviation using Equations (5) and (6).com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. the result being the difference between the readings for the analyte and the reference material.24). Some analysis processes appear at first sight to be quite complex.6 N3. These are all referred to as Type B evaluations because the values were not obtained by statistical means. taking into account any sensitivity coefficients (see paragraphs 3. Add any interdependent components algebraically (i. Manufacturers' specifications. Further detail on the means of evaluation is given in Sections 4 and 5.e.1 (a) (b) (c) (d) (e) (f) United Kingdom Accreditation Service.ukas. much of the process can be considered to be negatively correlated and the uncertainty of measurement can be evaluated from the resolution and repeatability of the process. combine them by taking the square root of the sum of the squares. In such cases. Process The process of assigning a value of uncertainty to a measurement result is summarised below: Identify all sources of error that are likely to have a significant effect.10 for Type A evaluations and in paragraph 3.7 N4 N4. N3. Consider each uncertainty component and decide whether any are interrelated and whether a dominant component exists (see D3 and Appendix C respectively). It is recognised that in certain areas of testing it may be known that a significant contribution to uncertainty exists but that the nature of the test precludes a rigorous evaluation of this contribution. and their relationship with the measurand. All other relevant information.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 (c) Experience with or general knowledge about the behaviour and properties of similar materials and equipment.4 N3. TW13 4UN Website: www.5 Definitions are given in paragraph 3. However the influence of random effects is often evaluated by the use of statistics. ISO/IEC 17025:2005 requires that a reasonable estimation be made and that the form of the reporting does not give an incorrect impression of the uncertainty.11 for Type B evaluations. dilutions and processing before results are obtained. in the absence of a dominant component.
2147 High Street. UKAS Copyright exists in all UKAS publications PAGE 77 OF 82 . If the nonrepeatability of the system is significant. In the case of a dominant component derived from Type B evaluation. selected on the basis of the coverage probability required. to provide the expanded uncertainty U (see paragraphs 3. such as are described in References [7] and [8].42 to 3.44). if required. (h) N4.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.3 United Kingdom Accreditation Service.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 (g) Multiply the combined standard uncertainty by a coverage factor k. coverage factor and coverage probability in accordance with Section 6. Report the result and. If one uncertainty contribution is significantly larger than the others then modifications may be required to this procedure. TW13 4UN Website: www. the expanded uncertainty. Middlesex. Further information regarding uncertainty evaluation for testing activities can be obtained from specialist publications that address particular fields of testing. Feltham.2 N4.ukas. and its effects are evaluated by using a Type A analysis. see Appendix C. it may be necessary to use the procedure in Appendix B.
Middlesex. The following data set is presented as an example: 1000. s. or are written down to a reasonable amount of significant figures.2 The value of s that is obtained from this data. This Appendix gives brief details of precautions that may be necessary under these circumstances. It is recommended that readers of this document gain familiarity with the functions involved by practicing the calculations presented in Appendix K .ukas. in order to prevent the cumulative effects of rounding errors having a significant effect on the result. and amending. Feltham. UKAS Copyright exists in all UKAS publications PAGE 78 OF 82 . Mention is also made of other techniques of uncertainty evaluation that electronic data processing has made practical. Most scientific calculators work with sufficient accuracy so that they do not in themselves introduce any significant errors .1 P4. where possible.3 1000.1 P2. should be 0. (a) (b) (c) United Kingdom Accreditation Service.2 P4 P4. Use of calculators Most scientific calculators are easily capable of all the calculations required for the evaluation of measurement uncertainty. The time spent constructing the spreadsheet can easily be recovered by the subsequent ease of producing. Calculation of standard deviations Many scientific calculators include statistical functions for calculation of standard deviations in accordance with Equation 5. Use of spreadsheets The widespread use of personal computers has made repetitive calculations a much easier process than in the past.1 Introduction Due to the nature and quantity of the calculations involved it is inevitable that some form of electronic processing will be involved in these calculations. or Display an incorrect nonzero value. It is recommended that a sample of results produced by a spreadsheet programme is checked manually using a scientific calculator to ensure that correct results are being generated.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX P ELECTRONIC DATA PROCESSING P1 P1.019 1000. using Equation 5. intermediate results are stored in the calculator memory for use later. sometimes. It is also recommended that. It will often be the case that this particular function is not capable of evaluating small values of standard deviation correctly. It is possible to construct a spreadsheet using the various equations presented in this document in order to perform the uncertainty calculations. uncertainty budgets.with one notable exception (see Section P4). Most calculators will either: Display an error message.2 P3 P3.004163.this can give rise to a better understanding of the process as well as giving the users confidence in their own abilities. 2147 High Street. TW13 4UN Website: www.025 P4. The calculator key associated with this function will usually be marked σ n1 or.1 P3.015 1000.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service.021 P2 P2. or Display a value of zero.
In a Monte Carlo simulation the mathematical model of the underlying system is run over and over again. 15. Such techniques are useful for validating the results returned by the application of the GUM. these form the distribution of the output.4 P5.5 It is possible that electronic spreadsheets will also suffer from similar errors therefore suitable checks should be devised to evaluate any such effects.2 P5. as well as in circumstances where the assumptions made by the GUM do not apply. From the output distribution confidence intervals can be produced. Other techniques The use of high speed data processing has opened new avenues for the analysis of uncertainty.025. A concept of likelihood ratio allows the ‘prior’ odds to be adjusted according to the each piece of new evidence. Bayes’ Theorem is particularly useful for analysis of qualitative data. Middlesex. provide an alternative approach to uncertainty evaluation in which the propagation of uncertainties is undertaken numerically rather than analytically. i. 19 and 21. making use of the ’prior’ odds that the belief is correct.3 P5. then.4 A solution to this problem is to use only the few least significant digits in the set of data. an estimate of the plausibility of the belief before one has the new data. the numbers 0. P5 P5. these techniques are able to provide much richer information. for the data above. Each run of the model is called a simulation. UKAS Copyright exists in all UKAS publications PAGE 79 OF 82 . 0.5 United Kingdom Accreditation Service.00416. or trial. So. P4. as can other statistical information. each time using a different set of random numbers representing the input variables. Bayes’ Theorem gives the ’posterior’ odds on the correctness of a belief (given the new evidence that has just been observed). Another method of uncertainty analysis is by the use of Bayesian statistics. by propagating the distributions (rather than just the uncertainties) of the inputs x i through the measurement model f to provide the distribution of the output y. this is based on the concept of degree of belief. One such technique is known as Monte Carlo simulation. the whole range of possible outputs can be produced. and at the end of each trial the outcome of the process is recorded. It can also be applied to the expression of professional opinions that may appear in test reports. 25. NOTE This could equally have been evaluated using just the last two digits. as this will then appear in the correct place in the results thereby minimising the likelihood of errors being made. by running a large enough number of trials (each with a different set of random numbers). Sampling techniques.e. 0.019 and 0. where the conventional GUM methodology is not appropriate.ukas.e. such as Monte Carlo simulation. Each of these sets of random numbers combines via the model to represent a different outcome. 2147 High Street. In fact. Feltham. i. and so on until all the evidence is used and a probability associated with the belief is obtained.1 P5. Each different outcome arises. However it is useful to include a decimal point. corresponding to a particular set of random numbers being applied to it.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. TW13 4UN Website: www.021 can be entered. If the model is a good representation of the realworld system. through the measurement model. yielding the correct value of s(qj): 0.015.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 P4. Another iteration is then performed using the next piece of evidence.
ci f ∂ xi f/∂ k kp ks m n N qj q p σ s(qj) s q tp(ν ) eff Experimental standard deviation of the mean value q . See Note 1 below.ukas.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. The meanings have also been described further in the text. Partial derivative with respect to input quantity Xi of the functional relationship f between the measurand and the input quantities.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX Q SYMBOLS The symbols used are taken mainly from the GUM. Estimate of the standard deviation σof the population of values of a random variable q based on a limited sample of results from that population. TW13 4UN Website: www. but are summarised here for convenience of reference. Number of readings or observations that contribute to a mean value. UKAS Copyright exists in all UKAS publications PAGE 80 OF 82 . The standard deviation of a population of data using all the samples in that population.. Functional relationship between the measurand Y and the input quantities X i on which Y depends. Middlesex. Sensitivity coefficient used to multiply the value x i of an input quantity X i to express it in terms of the measurand Y. See Note 1 overleaf. probability distributions unknown. N. Coverage factor (general).. where i = 1 . and between output estimate y and input estimates xi upon which y depends. Feltham. Arithmetic mean or average of n repeated observations of randomly varying quantity q. 2147 High Street. Coverage factor used to calculate an expanded uncertainty Up for a specified coverage probability p. usually where they first occur. jth repeated observation of randomly varying quantity q. ai ad Estimated semirange of uncorrelated systematic component of uncertainty.. Coverage probability or level of confidence expressed in percentage terms or in the range zero to one. A systematic component of uncertainty that so dominates other contributions to uncertainty in magnitude that special consideration has to be given to its presence in calculating the expanded uncertainty.. Student tfactor for ν degrees of freedom corresponding to a given coverage probability p. if different from n. eff United Kingdom Accreditation Service. Number of readings or observations that are used for the evaluation of s(q j). Number of input estimates xi on which the value of the measurand depends. Coverage factor chosen for the purpose of comparison with a specification limit.
TW13 4UN Website: www.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. UKAS Copyright exists in all UKAS publications PAGE 81 OF 82 . Up ν ν i ν eff NOTE 1 United Kingdom Accreditation Service. Feltham. Degrees of freedom. Combined standard uncertainty of output estimate y. M3003 uses the subscript j instead of k in order to avoid any possible confusion with the coverage factor k. in general. Effective degrees of freedom of u c(y) used to obtain tp(ν ) eff The GUM uses the symbols qk and s(qk ) where qj and s(qj) are used here. 2147 High Street. Expanded uncertainty of output estimate y that describes the measurand as an interval Y = y ± U. Degrees of freedom of standard uncertainty u(xi) of input estimate xi .THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 u(xi) u c(y) U Standard uncertainty of input estimate xi . Expanded uncertainty of output estimate y that describes the measurand as an interval Y = y ± Up. Middlesex. with a high coverage probability.ukas. with a specified coverage probability p. the number of terms in a sum minus the number of constraints on the terms of the sum.
ANAMET Connector Guide. ISBN 0 948926 15 5 3 4 5 6 7 8 United Kingdom Accreditation Service. OIML. Institute of Physics. International Organisation for Standardization. Guide to the Expression of Uncertainty in Measurement. United Kingdom Accreditation Service. Reexamination of mismatch uncertainty when measuring microwave power and attenuation.A and WARNER. Switzerland. IFCC. United Kingdom Accreditation Service. Uncertainty. F. IEC. IUPAC.com Publication requests Tel: +44 (0) 20 8917 8421 Fax: +44 (0) 20 8917 8500 © United Kingdom Accreditation Service. LAB34 Edition1. December 1999 HARRIS. The Expression of Uncertainty in EMC Testing.L. ISO. 2147 High Street. Part 3. UKAS Copyright exists in all UKAS publications PAGE 82 OF 82 . EA4/02 Expression of the Uncertainty of Measurement in Calibration. Second Edition. November 2006. Guide to the Expression of Uncertainty in Measurement.F. IEE Proceedings.ukas. 2000. LAB14 Edition 4. Calibration and Probability: The Statistics of Scientific and Industrial Measurement Uncertainty. First Edition 1993.1.THE EXPRESSION OF UNCERTAINTY AND CONFIDENCE IN MEASUREMENT M3003  EDITION 2  JANUARY 2007 APPENDIX R REFERENCES 1 BIPM. Calibration of Weighing Machines. February 1981. Edition 2. I. National Physical Laboratory. DIETRICH. BSI Equivalent: BSI PD 6461: 1995. January 1991. Geneva. August 2002. IUPAP. TW13 4UN Website: www. Vol 128 Pt H No 1. Vocabulary of metrology. BSI ISBN 0 580 23482 7 2 European cooperation for Accreditation. ISBN 9267101889. Feltham. QUAM:2000. Middlesex. C. EURACHEM/CITAC Guide: Quantifying Uncertainty in Analytical Measurement.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.