Chap 10 Sampling

Page 421
CHAPTER 10
Audit sampling
Copyright © 2018. McGraw-Hill Australia. All rights reserved.
Gay, Grant E., and Roger Simnett. Auditing and Assurance Services in Australia, McGraw-Hill Australia, 2018. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/adelaide/detail.action?docID=5729228.
Created from adelaide on 2021-05-12 00:46:38.
LEARNING OBJECTIVES (LO)
Note: For ease of understanding, we have divided this chapter into two
sections, the first being the theory of sampling (learning objectives 10.1–10.6),
the second a more detailed consideration of current sampling techniques
(learning objectives 10.7–10.9).
Sampling theory
10.1 Define audit sampling and its objective and describe the
requirements that apply to all audit samples—statistical and non-
statistical.
10.2 Identify the various means of gathering audit evidence.
10.3 Identify planning and design considerations for sampling,
including defining audit objectives and the appropriate
population, the potential use of stratification and potential
alternative definitions of the sampling unit, including dollar-unit
sampling.
10.4 Identify factors influencing the determination of sample size.
10.5 Understand the appropriate methods of selecting sample items.
10.6 Appreciate the application of audit procedures to a selected
sample, and the evaluation of sample results.
Sampling techniques
10.7 Consider sampling approaches to tests of controls, in particular,

attribute sampling.
10.8 Consider sampling approaches to substantive tests, in particular,
dollar-unit sampling.
10.9 Consider other statistical sampling approaches to substantive
tests.
RELEVANT GUIDANCE
ASA 500/ISA 500 Audit Evidence
ASA 530/ISA 530 Audit Sampling
Page 422
CHAPTER OUTLINE
While previous chapters have discussed the types of audit testing (including
tests of controls and substantive tests of transactions and balances) that the
auditor may undertake, at this stage it is still unclear how many transactions or
account balances an auditor should test, which specific transactions or account
balances, and how the auditor will know that they have collected sufficient
appropriate evidence to support their conclusion. These issues will be
addressed in this chapter. Audit sampling is the application of audit procedures
to less than 100 per cent of the items within a population. The objective is to
provide a reasonable basis for the auditor to draw conclusions about the
population from which the sample is selected (ASA 530.4/ISA 530.4). This
chapter discusses the steps to follow when undertaking audit sampling,
including determining the extent of audit tests, selection of sample items and
evaluation of sample results. How this chapter fits into the overall financial
report audit is illustrated in Figure 10.1 , which is an expansion of part of the
overall flowchart provided in Chapter 1 .
FIGURE 10.1 Flowchart of audit sampling for a financial report audit
Page 423
LO 10.1 Definition and features

Audit sampling is the application of an audit procedure to less than 100 per cent of the
items within a population to obtain audit evidence about particular characteristics of
the population. By undertaking audit tests on items included in the sample, the auditor will
have a reasonable basis on which to draw conclusions about the population from which the
sample is selected.
Some auditing procedures can be applied by sampling, but many auditing procedures do
not involve sampling. For example, enquiry, observation, analytical procedures and general
procedures, such as reading minutes and contracts, do not involve sampling. Also, where
the auditor has the ability to apply advanced data analytics to a population, or otherwise
test a population, sampling procedures are not applied. Further, sampling is not involved
when the auditor applies a procedure only to the significant items in an account balance or
reviews a client’s completed accounting routines.
While Chapters 8 and 9 have provided a distinction between different types of audit
evidence and the circumstances in which each type will be sought, audit sampling is
important because it provides information on:
how many items to examine

which items to select
how sample results are evaluated and extrapolated to the population.
Two features of particular importance to understanding audit sampling are sampling risk
and characteristic of interest.
Sampling risk
Sampling risk is the probability that the auditor has reached an incorrect conclusion
because audit sampling was used rather than 100 per cent examination (ASA 530.5/ISA
530.5). While sampling risk can be reduced to an acceptably low level by using an
appropriate sample size and selection method, it can never be eliminated. In other words,
the only way to eliminate sampling risk is to apply the audit procedure to every item
instead of to a sample. There are two types of sampling risk:
1. The risk that the auditor will conclude, in the case of a test of controls, that control risk is
higher than it actually is, or, in the case of a substantive test, that a material error exists
when in fact it does not. Audit efficiency is affected by this type of risk, because it would
usually lead to additional work to establish that initial conclusions were incorrect.
2. The risk that the auditor will conclude, in the case of a test of controls, that control risk is
lower than it actually is (that controls are more effective than they actually are), or, in the
case of a substantive test, that a material error does not exist when in fact it does. Audit
effectiveness is affected by this type of risk, as it is more likely to lead to an erroneous
auditor’s opinion. It is therefore of greater concern to the auditor.
Sampling risk can be compared with non-sampling risk , which arises from factors
other than sample size that cause the auditor to reach an incorrect conclusion. Such factors
include the possibility that the auditor will fail to recognise misstatements contained in
examined items, or that the auditor will apply a procedure that is not effective in achieving
a specific objective. Generally, auditors try to hold this non-sampling risk to a negligible
level through adequate planning and supervision and appropriate quality control.
Characteristic of interest
The second feature of the definition of audit sampling distinguishes sampling from other
approaches to audit tests. The auditor examines a sample of items to obtain evidence about
particular characteristics of the population.
When the auditor uses sampling in tests of controls, the characteristic of interest is
usually a deviation rate from an internal control policy or procedure. In other words, the
auditor is concerned with the rate of deviation from a prescribed policy or Page 424
procedure, such as the matching of a supplier’s invoice with a goods received note
before payment is authorised, and specifically with whether the deviation rate exceeds a
rate the auditor has specified as acceptable.
When the auditor uses sampling in substantive tests, the characteristic of interest is
monetary misstatement in the balance. In other words, the auditor is concerned with the
dollar amount of misstatement and specifically with whether the monetary misstatement
exceeds a material amount.
QUICK REVIEW
1. Audit sampling is the application of an audit procedure to less than 100 per
cent of the items in a population, to obtain audit evidence about a
particular characteristic of the population.
2. Audit sampling is important because it provides information on:
how many items to examine
which items to select
how sample results are evaluated and extrapolated to the whole
population.
3. Sampling risk is the probability that the auditor reached an incorrect
conclusion because audit sampling was used rather than 100 per cent
examination.
4. When sampling, the major characteristics of interest are monetary
misstatement of an account balance or transaction class (substantive
tests), and deviation rates for an internal control (tests of controls).
LO 10.2 Various means of gathering audit evidence
ASA 500 (ISA 500) discusses the need to collect and evaluate sufficient and appropriate
audit evidence. The means available to the auditor for selecting items for testing are as
follows (ASA 500.A52/ISA 500.A52):
1. 100 per cent examination While examination of every item may be an appropriate
evidence-gathering technique in certain situations, it is not a sampling method. It means
examining the entire population (100 per cent) of items that make up an account balance
or class of transactions (or a subgroup within that population). Advanced data analytics,
such as those discussed in Chapter 7 , and the concept of big data have meant that this
evidence gathering technique is more likely to be used, but it does not mean that
sampling is not applicable. Even in circumstances where advanced data analytics are
capable of analysing large populations, and identifying many transactions or account
balances that need the auditor’s attention, the auditor may have to sample from these
risky items and use sampling techniques such as those outlined in this chapter to be able
to reach a conclusion on the extent of misstatement in a population. Further, 100 per cent
examination techniques are commonly used when the population constitutes a small
number of large-value items, and when both inherent and control risks are evaluated as
high and other means do not provide sufficient appropriate audit evidence.
2. Selecting specific items Similarly, selecting specific items in a population is not a
sampling method. The auditor may decide to select specific items within a population
because they are of high value or exhibit some other characteristic of interest, for example
if they are suspicious, unusual, particularly risk-prone or have a history of error. A
common technique in practice is to select all items over a specific dollar value or of a
particularly risky nature in order that the auditor concentrates on the more risky and
material items. The results of audit procedures applied to items selected in this way
cannot be projected to the entire population because not all items in the population had a
chance of being selected.
3. Audit sampling This means that all sampling units in the population should have an
opportunity (preferably an equal opportunity) of being selected. If this is achieved, it
gives the best chance for the sample to be representative of the population. Audit
sampling is a common evidence-gathering technique, and can involve either a statistical
or a non-statistical approach.
Page 425
10.1 Auditing in the global news ...
How big data and advanced data analytics are transforming the
audit
‘… recent technology advancements in big data and analytics are providing
an opportunity to rethink the way in which an audit is executed.
The transformed audit will expand beyond sample-based testing to include

analysis of entire populations of audit-relevant data (transaction activity and
master data from key business processes), using intelligent analytics to
deliver a higher quality of audit evidence and more relevant business
insights. Big data and analytics are enabling auditors to better identify
financial reporting, fraud and operational business risks and tailor their
approach to deliver a more relevant audit.
While we are making significant progress and are beginning to see the
benefits of big data and analytics in the audit, we recognize that this is a
journey.’
Source: EY (2015) ‘How big data and analytics are transforming the audit’, www.ey.com/gl/en/service
s/assurance/ey-reporting-issue-9-how-big-data-and-analytics-are-transforming-the-audit#item.
Non-statistical versus statistical audit sampling

When an auditor uses audit sampling, the basic principles and essential procedures
identified in ASA 530 (ISA 530) apply equally to both statistical and non-statistical
approaches.
Statistical sampling is defined in ASA 530.5(g) (ISA 530.5(g)) as:

an approach to sampling that has the following characteristics:
(a) random selection of the sample items; and

(b) the use of probability theory to evaluate sample results,
including measurement of sampling risk.
Non-statistical sampling is therefore any sampling approach that does not have both of
the characteristics of statistical sampling.
Both of the two essential characteristics contained in the definition of statistical sampling
(random sample selection and probability theory) must be met for the sample to qualify as
statistical. For example, use of a random number table to select sample items does not
mean that statistical sampling is being used unless the sample results are evaluated
mathematically. Mathematical evaluation based on a non-random sample is not valid.
The mathematical evaluation that is the distinguishing feature of statistical sampling leads
to quantification of sampling risk. One of the advantages of statistical sampling is that
sampling risk can be objectively calculated as a percentage and controlled precisely by
adjusting sample size. An auditor using non-statistical sampling has to consider sampling
risk and hold it to an acceptable level, but cannot quantify sampling risk precisely. It is
argued that the benefits of quantification of sampling risk are that it allows the auditor to
explain and defend the decisions undertaken in the audit.
Distinction between statistical and non-statistical sampling
There is clearly a similarity between the essential features and basic requirements
applicable to all audit samples, whether statistical or non-statistical. The requirement for
opportunity of selection that applies to all audit samples is achieved by random selection.
In practice, however, some selection methods are considered to produce a representative
sample although they do not qualify as random selection, and should not be used Page 426
for statistical sampling. The requirement for projection of sample results that
applies to all audit samples is a form of mathematical evaluation. Therefore, the chief
conceptual difference between the two types of sampling is that the mathematical
evaluation of a statistical sample includes quantification of sampling risk, as outlined
earlier. For both statistical and non-statistical sampling, sample size may be determined by
the auditor’s judgment.
The choice of statistical or non-statistical sampling
The decision whether to use a statistical or non-statistical approach is a matter for the
auditor’s judgment about the most efficient means of obtaining sufficient appropriate audit
evidence in the particular circumstances. For example, when performing tests of controls,
the auditor’s analysis of the qualitative characteristics of deviations (for example,
considering the reasons why they occur, and whether similar deviations are likely to occur)
is often more important than the mere presence or absence of deviations. On many
occasions, the best ways of incorporating the impact of qualitative characteristics of
deviations are through utilising the auditor’s professional judgment and experience.
Another reason some audit firms prefer non-statistical to statistical sampling is that
deviations or misstatements in a population are not random. For example, computer
environments lead to repetitive errors; many fraud schemes are systematic; and
management may override only some controls. However, in these situations, unless
auditors know something about the expected pattern of deviations or irregularities and can
use this knowledge in developing their sampling plan, there is no reason why non-statistical
sampling approaches should be preferred to statistical sampling approaches.
Current practice in Australia and the impact of the business

risk assessment
In Australia, the larger auditing firms are exploring, and utilising for some clients, data
analytics. This is expected to grow significantly over the next few years. With regard to the
practice of sampling, there are some disparities within the large audit firms. Firms will
usually use an unbiased approach such as random sampling to select their sample items, but
the size of the sample they select will usually be determined with the help of decision aids
within the firm. Sample sizes that are commonly used in practice are around 20 items
where a moderate level of testing is required, or 30 items where more extensive levels of
testing are undertaken.
The sample sizes used in practice are commonly smaller than those computed using the
techniques for determining sample size outlined later in this book. There are two main
reasons for this. The first is that a reduction in emphasis on undertaking large samples is
supported by the business risk approach, which emphasises a top-down approach to the
audit whereby an understanding is gained of the entity’s business, strategy and associated
risks. Large sampling applications are more consistent with a bottom-up approach whereby
characteristics of a population are inferred from examining specific transactions or account

balances. The second is that the auditor rarely gains all their confidence about the level of
misstatements from only one type of audit testing. This emphasises the cumulative nature
of audit evidence (ASA 500.A1/ISA 500.A1), with multiple types of evidence being
obtained during the course of the audit which can help the auditor assess the different risks
of misstatement.
Basic requirements of all audit samples
Whenever an auditor uses audit sampling (statistical or non-statistical) the following
requirements, illustrated in Figure 10.2 , apply:
FIGURE 10.2 Sampling phases
Page 427
Planning and design In planning and designing an audit sample, the auditor
considers the relationship of the sample to the relevant specific audit objective or control
objective and considers certain other factors that should influence sample size. (These
other factors are explained later.)
Selection Sample items are selected in such a way that the sample can be expected to be
representative of the population. Thus, all items in the population have an opportunity of
being selected. (The methods that meet this requirement for representative selection are
explained later.)
Performing the procedure and evaluating the results The auditor performs the
required audit procedures on the items selected, projects the results of the audit
procedures undertaken on the sample to the population and considers sampling risk. (The
methods for projecting sample results are explained later. The sample results would be a
monetary misstatement for a substantive test and deviation rate for tests of controls.)
QUICK REVIEW
1. The auditor might decide to select specific items within a population
because they are of high value or because they are assessed as higher
risk. This is not audit sampling, as the results of audit procedures applied
to items selected in this way cannot be projected to the entire population
because not all items in the population had a chance of being selected.
2. Audit sampling can be either statistical or non-statistical. Statistical
sampling involves the use of (a) random sample selection and (b)
probability theory to evaluate sample results, including measuring
sampling risk. Non-statistical sampling is any sampling approach that does
not have both the characteristics of statistical sampling.
3. The business risk approach and cumulative nature of audit evidence
means that sample sizes used in practice are commonly smaller than
those computed to reduce sampling risk to a desired level.
LO 10.3 Planning and designing the sample
The auditor should design a sample carefully to ensure that it achieves the required audit
objective. The auditor needs to consider:
the objectives of the audit test

the population from which to sample
the possible use of stratification
the definition of the sampling unit.
Audit objectives
The auditor first considers the specific audit objectives and the combination of audit
procedures that is likely to best achieve them (ASA 530.A5/ISA 530.A5). Consideration of
the nature of the audit evidence sought and possible deviation conditions or other
characteristics relating to that evidence assist the auditor in defining what constitutes a
deviation or misstatement and what population to use for sampling.
The auditor considers what conditions would constitute a deviation or misstatement (ASA
530.A6/ISA 530.A6). For example, in testing the control of approval of purchase
requisitions, the auditor needs to know who may approve such requisitions and how this
approval would be evidenced. Thus missing signatures or approvals by persons not duly
authorised constitute a deviation. A clear understanding of what constitutes a deviation is
important to ensure that all those conditions that are relevant to the audit objectives, and
only those, are included in the projection of errors. For example, in a substantive procedure
aimed at verifying the existence of accounts receivable, such as confirmation, payments
made by the customer before the confirmation date but received shortly after that date by
the client are a timing difference rather than an error. Also, an error of posting to an
incorrect customer account does not affect the total accounts receivable balance. Page 428
Therefore it is not appropriate to consider such situations as misstatements when
evaluating the sample results.
Population
The population is the entire set of data about which the auditor wishes to draw conclusions
and from which a sample is selected. It is therefore essential for the auditor to ensure that
the population identified is both complete and appropriate to the objective of the sampling
procedure.
Completeness For example, if the auditor intends to undertake audit tests on the accounts
receivable master file, the first thing the auditor will need to do is ensure that the total of
the accounts receivable master file from which the auditor is going to sample is
reconciled with the accounts receivable balance in the general ledger. Similarly, if the
auditor intends to use the sample to draw conclusions about the operation of an internal
control during the financial reporting period, the population must include all instances
when this control should be initiated.
Appropriateness This includes consideration of the direction of testing. For example, if
the auditor’s objective is to test for existence of accounts payable, the population could be
defined as the accounts payable listing. On the other hand, when testing for completeness
of accounts payable, the population is not the accounts payable listing but rather
subsequent disbursements, unpaid invoices, suppliers’ statements, unmatched receiving
reports or other populations that by testing will provide audit evidence of the non-
recording of accounts payable.
Stratification
Audit efficiency may be improved if the auditor stratifies a population. This means
dividing it into discrete sub-populations (strata) which have an identifying characteristic,
such as dollar value (ASA 530/ISA 530 Appendix 1, paragraphs 1–4).
When performing substantive procedures, an account balance or class of transactions is

often stratified by dollar value, allowing greater audit effort to be directed to the large-
value items, which contain the greatest potential for monetary error in terms of
overstatement. An example of this is contained in Global example 10.1 . Similarly, a
population may be stratified by risk of error: for example, when testing the
accuracy, valuation and allocation assertion of accounts receivable, debtors may be
stratified by age.
GLOBAL EXAMPLE 10.1 Stratification of accounts receivable
In testing the accounts receivable master file, the auditor may identify three
strata: accounts greater than $1 million, accounts between $10 000 and $1
million, and accounts less than $10 000. Stratification reduces the variability
of items within each stratum and therefore allows sample size to be reduced
without a proportional increase in sampling risk. Sub-populations need to be
carefully defined such that any sampling unit can belong only to one stratum.
The auditor may then test 100 per cent of the accounts greater than $1
million (note: this is not a sampling approach, as it involves testing 100 per
cent of the items in this stratum), 10 per cent of the items for the strata $10
000 to $1 million and 2 per cent of items less than $10 000.
It should be noted, however, that the results of procedures applied to a sample of items
within a stratum can be projected only to the items that make up that stratum. To conclude
on the entire population, the auditor needs to consider the impact of potential misstatement
in other strata. For example, 20 per cent of the items in a population may comprise 90 per
cent of the value of an account balance. The auditor may decide to examine a sample of
these items. The auditor would evaluate the results of this sample and conclude on the 90
per cent of the value separately from the remaining 10 per cent (which may be either
considered immaterial, or for which a further sample or other means of gathering evidence
would be used).
Page 429
Defining the sampling unit

The sampling unit is commonly each of the transactions (especially for tests of controls
where controls, such as authorisations, are related to transactions) or balances making up
the account balance (for example, if sampling from accounts receivable, each individual
debtor is a sampling unit).
A sampling technique specific to auditing is a technique called value-weighted selection

(ASA 530/ISA 530 Appendix 1, paragraph 5). It is also commonly referred to as monetary-
unit sampling (MUS) or, in countries where the currency units are dollars, dollar-unit sam
pling (DUS) . Under this technique the sampling units are individual dollars rather than
each transaction or balance. It is commonly used in auditing because it is often efficient in
substantive testing for the auditor to identify the sampling unit as the individual dollars that
make up an account balance or class of transactions. This approach to defining the
sampling unit ensures that audit effort is directed to the large-value items because they
have a greater chance of selection, and has the advantage of resulting in smaller sample
sizes than would be required under traditional sampling approaches where the individual
transactions or components of account balances (such as individual debtors or inventory
items) comprise the sampling units. Thus, a balance containing $10 000 will be 10 times
more likely to be selected than an account containing $1000. Having selected individual
dollars from within the population of, for example, the accounts receivable balance, the
auditor then examines the particular items, for example individual debtors, that contain
those dollars. DUS will be further discussed later in this chapter.
QUICK REVIEW
1. In planning and designing the sample, the auditor carefully considers the
objectives of the audit procedure. This will determine the attributes of the
population from which the sample is to be drawn. The auditor should
ensure that the population is complete and appropriate for the objectives
of the sampling procedure.
2. Audit efficiency may be improved if the auditor stratifies a population,
which involves dividing the population into discrete sub-populations which
have an identifying characteristic such as dollar value. As a result of
reduced variability within each stratum, sample size can be reduced
without an increase in sampling risk.
3. Sampling units in auditing are usually each transaction, component of
account balance or each individual dollar making up an account balance
or class of transactions.
LO 10.4 Determining sample size
The auditor’s major consideration in determining the appropriate sample size is whether
sampling risk will be reduced to an acceptably low level (ASA 530.7/ISA 530.7). Sample
size is affected by the degree of sampling risk that the auditor is willing to accept. As
outlined earlier, sampling risk is the probability that the auditor has reached an incorrect
conclusion because sampling, rather than 100 per cent examination, was used. While
sampling risk can be reduced to an acceptably low level by using an appropriate sample
size and selection method, it can never be eliminated when sampling is being used. The
lower the risk the auditor will accept, the greater the sample size needs to be.
The sample size can be determined by the application of a statistical formula, through the
use of decision aids available to auditing firms or through professional judgment applied to
the circumstances. When a formula is used, the auditor documents the reasons for selecting
the various values used in the formula. Similarly, when judgment alone is used, the auditor
documents the factors considered and their impact on the sample size determined.
Appendix 2 to ASA 530 (ISA 530) outlines the factors that influence sample size Page 430
for tests of controls (see Exhibit 10.1 ).
FACTORS THAT INFLUENCE SAMPLE SIZE FOR TESTS
EXHIBIT 10.1
OF CONTROLS
1. The extent to which the auditor’s risk assessment takes into account relevant
controls (control risk assessment) The more assurance the auditor intends to
obtain from internal control, the lower the assessment of control risk and the larger
the sample size needs to be. For example, a preliminary assessment of control risk
as low indicates that the auditor plans to place considerable reliance on the effective
operation of particular internal controls. The auditor therefore needs to gather more
audit evidence to support this assessment than would be the case if control risk
were assessed at a higher level (that is, if less reliance were planned).
2. The tolerable rate of deviation The lower the rate of deviation from a control
that the auditor will accept, the larger the sample size needs to be.
3. The expected rate of deviation in the population The higher the expected
rate of deviation from a control, the larger the sample size needs to be, so that the
auditor is in a position to make a reasonable estimate of the actual rate of deviation.
Factors relevant to this consideration include the auditor’s understanding of the
business (in particular, procedures undertaken to obtain an understanding of the
internal control), changes in personnel or in internal control, the results of audit
procedures applied in previous periods and the results of other audit procedures.
High expected deviation rates from a control usually mean that tests of controls are
omitted, and the auditor increases substantive testing.
4. The auditor’s desired level of assurance that the actual deviation rate is not
larger than the tolerable deviation rate The greater the level of confidence that
the auditor requires that the results of the sample are in fact indicative of the actual
incidence of deviation in the population, the larger the sample size needs to be.
5. The number of sampling units in the population For larger populations,
commonly specified as greater than 3000 items, the actual size of the population
has little effect on sample size.
Source: (c) 2018 Auditing and Assurance Standards Board (AUASB). The text, graphics and layout of this
publication are protected by Australian copyright law and the comparable law of other countries. No part
of the publication may be reproduced, stored or transmitted in any form or by any means without the
prior written permission of the AUASB except as permitted by law. For reproduction or publication
permission should be sought in writing from the Auditing and Assurance Standards Board. Requests in
the first instance should be addressed to the Technical Director, Auditing and Assurance Standards
Board, PO Box 204, Collins Street West, Melbourne, Victoria, 8007.
Appendix 3 to ASA 530 (ISA 530) outlines the factors that influence sample size for
substantive testing (see Exhibit 10.2 ).
FACTORS THAT INFLUENCE SAMPLE SIZE FOR
EXHIBIT 10.2
SUBSTANTIVE TESTING
1. The auditor’s assessment of the risk of material misstatement This is the

auditor’s combined assessment of inherent and control risk, as described in
Chapter 4 . A higher inherent risk means that auditors will have to collect more
audit evidence in order to reduce audit risk to an acceptable level. A higher control
risk means that the auditor cannot place as much reliance on the effective operation
of internal controls to reduce the risk of material misstatement and will therefore
need to gather further evidence by substantive testing. If the further audit evidence
involves sampling, then this will be achieved by an increase in sample size.
2. The use of other substantive procedures directed at the same assertion The
more the auditor relies on other substantive procedures (tests of details or analytical
procedures) to reduce the detection risk regarding a particular account balance or
class of transactions, the less the assurance required from sampling and, therefore,
the smaller the sample size can be.
3. The auditor’s desired level of assurance that actual misstatement is Page 431
not greater than tolerable misstatement The greater the confidence
required by the auditor that the results of the sample are indicative of the actual
misstatement in the population, the larger the sample size needs to be. For example,
the auditor may require greater confidence in areas where management has greater
discretion, such as the expense versus capitalise decision for repairs and
maintenance, thus the auditor may have to look at a larger sample for these items.
4. The tolerable misstatement The lower the auditor’s tolerance for
misstatement in the population, the larger the sample size needs to be. For
substantive procedures, the auditor’s tolerable misstatement will be the same
amount or an amount lower than performance materiality, discussed in
Chapter 4 (ASA 530.A3/ISA 530.A3).
5. The amount of misstatement the auditor expects to find in the population (exp
ected misstatement ) The higher the expected misstatement, the larger the
sample size needs to be so that the auditor is in a position to make a reasonable
estimate of the actual misstatement. Factors relevant to the auditor’s consideration
of the expected misstatement amount include the extent to which item values are
determined subjectively or by the application of complex formulas, the results of
tests of controls, the results of audit procedures applied in previous periods and the
results of other substantive procedures.
6. Stratification When the population has been stratified, each sub-population is
more homogeneous, and therefore the aggregate of the sample sizes from the sub-
populations is less than the sample size that would have been required had one
sample been drawn from the whole population.
7. The number of sampling units in the population For larger populations,
commonly specified as greater than 3000 items, the actual size of the population
has little effect on sample size.
Source: (c) 2018 Auditing and Assurance Standards Board (AUASB). The text, graphics and layout of this
publication are protected by Australian copyright law and the comparable law of other countries. No part
of the publication may be reproduced, stored or transmitted in any form or by any means without the
prior written permission of the AUASB except as permitted by law. For reproduction or publication
permission should be sought in writing from the Auditing and Assurance Standards Board. Requests in
the first instance should be addressed to the Technical Director, Auditing and Assurance Standards
Board, PO Box 204, Collins Street West, Melbourne, Victoria, 8007.
QUICK REVIEW
1. In determining the sample size the auditor should consider whether
sampling risk will be reduced to an acceptably low level.
2. Factors that will have a significant effect on sample size for tests of
controls are:
the auditor’s intended reliance on internal control
the tolerable rate of deviation
the expected rate of deviation
the desired level of assurance that the actual deviation rate is not larger
than the tolerable deviation rate.
3. Factors that will influence sample size for substantive testing are:
the auditor’s assessment of the risk of material misstatement
other substantive procedures directed at the same assertion
the desired level of assurance
the tolerable misstatement
the expected misstatement
stratification.
LO 10.5 Selecting the sample
Since the purpose of sampling is to draw conclusions about the entire population or stratum
from which the sample has been drawn, the sample needs to be typical of the
characteristics of that population or stratum. To achieve this, the sample needs to be
selected without bias, so that all sampling units in the population or stratum have a chance
of selection.
Appendix 4 of ASA 530 (ISA 530) outlines selection procedures. Sample items should be
selected in such a way that the sample is expected to be representative of the population.
As described in ASA 530.A13 (ISA 530.A13) there are three principal sampling Page 432
methods that achieve this: random selection, systematic selection and haphazard
selection.
Random selection
In random selection , each sampling unit making up the account balance or class of
transactions has a chance, often an equal chance, of selection. The concept requires that the
person selecting the sample will not influence or bias the selection either consciously or
unconsciously, so some form of impartial selection process is used to make the sample
truly random. Auditors commonly use computerised random number generators or random
number tables to generate numbers which can be linked to pre-numbered sampling units in
the population. If selecting from pre-numbered transactions the auditor can identify the
first and last transaction numbers for the period under audit, between which the random
numbers are generated. If sampling from components of account balances, such as
individual debtors or inventory items, the software can randomly select from the population
of valid debtor numbers or inventory item numbers.
This approach is appropriate for both non-statistical and statistical sampling, since the
sample is selected on a basis that allows the auditor to measure the probability of selecting
the combination of sampling units actually chosen.
Systematic selection
Systematic selection involves selecting every nth item in the population, the sample
interval being determined by dividing the number of items in the population by the sample
size, and selecting a random starting point. For example, if the auditor wants a sample of
20 from a population of 20 000 items, the sample interval is every 1000th item. The auditor
randomly selects a starting point (falling between the first item in the population and the
sample interval) and selects every 1000th item from that starting point.
This method produces a random sample only if the population is randomly arranged. For
example, if large-value sales invoices are always assigned a number in a particular group, a
systematic sampling process may result in a biased sample of primarily large or small
invoices. Again, software can be used to select systematic samples from computer-based
files. This method is illustrated by Global example 10.2 .
GLOBAL EXAMPLE 10.2 Sampling by systematic selection
Assume the sample size has been determined as 20 and the number of
items in the population is 10 000.
Step 1: Calculate the sample interval:

No. of items in population 10 000
= = 500
Sample size 20
Step 2: Give every item in the population an equal chance of selection by

choosing a random number (random start) within the range of 1 and the
sampling interval (in this example, 500), for example, 217.
Step 3: Continue to add the sampling interval to the random start to
identify sample items, e.g. 217, 717, 1217, 1717 . . . 9217, 9717.
This gives an even spread of items throughout the population.
Random selection and systematic selection are the most common sampling techniques. In
comparing these techniques, systematic selection may be advantageous in ensuring an even
distribution of selection throughout the population (for example, in testing the Page 433
continuity objective of internal control). However, if there are potential patterns in
the population (for example, control fails for every 100th transaction), care must be taken
in using systematic sampling: if the pattern corresponds to the sample interval an
unrepresentative sample will result.
Haphazard selection
Haphazard selection is permitted by the auditing standards, although it cannot
normally be used for statistical sampling applications because it does not allow the auditor
to measure the probability of selecting the combination of sampling units and it cannot
guarantee that the auditor will select without bias. In this method the auditor selects
sampling units without any conscious bias, and in a manner that can be expected to be
representative of the population. This technique may be useful when selecting a sample
from a population which is physically stored in an unsystematic manner. For example,
where all sales invoices are filed in a filing cabinet drawer, the auditor selects invoices
from the drawer irrespective of any distinctive feature of any of the individual documents.
Humans are, however, subject to subconscious biases which may affect the randomness of
this approach. These include an affinity with the number seven, and also a bias to select
towards the middle of a sequence rather than at the extremes. Both of these biases have
been demonstrated a number of times by research, with the affinity for the number seven
illustrated in Auditing in the global news 10.2 .
10.2 Auditing in the global news ...
Biases from haphazard sampling

This note describes a technique that we have used for several years to
illustrate the problems inherent in judgment sampling techniques.
The question (asked) . . . select a number from 1 to 10 and write it in the

space below . . . The results reported below are from a group of
approximately 140 students.
A truly random selection would have led to each number being selected by
approximately one-tenth of the class. This question is, however, intended to
highlight the human affinity for the number seven (our weeks are seven days
long, we split our oceans into seven seas, we are tempted to indulge in
seven deadly sins, Snow White had seven dwarfs, we drink 7-Up™ and so on).
Whether the students were aware of this or not, their selection of a ‘random’
number was hopelessly biased by this innate affinity.
Source: Stewart and Dunn, 2000.

Page 434
Unacceptable sample selection methods

In the past, auditors have used some selection methods that cannot be expected to produce
representative samples. While such methods of identifying items to test may be acceptable
evidence-gathering techniques, they are rarely suitable as an approach to selecting sample
items when the intention is to extrapolate the results from testing the sample to the
population. Examples are as follows:
Block selection With this method the auditor selects a block of items from a
population. For example, the auditor might examine all cash disbursement transactions in
the first week of June and the last week of December. The problem with this selection
method is that the sampling unit is a period of time rather than an individual transaction.
In the example given, the auditor has selected two units in a population of 52 units. If the
blocks or periods were selected randomly the sample might be representative, but a valid
sample size would normally be impractically large. For example, to have reasonable
assurance the auditor might need to select 30 out of 52 weeks.
Judgmental selection based on sample item characteristics With this method the
auditor selects large or unusual items from the population or uses some other judgmental
criterion for selection, with the result that there is bias towards high-value or key items.
This method is an example of ‘selecting specific items’, which was discussed earlier in
this chapter under Various means of gathering audit evidence , and is not a sampling
technique. Obviously, this method has a conscious bias and the items selected cannot be
considered representative of the population. Because these judgmental selection methods
have traditionally been used extensively in practice as a method of selecting sample
items, it is important to recognise that they are unacceptable today even for non-statistical
sampling.
The fact that judgmental selection with bias towards high-value or key items is not
appropriate for audit sampling does not mean that auditors should stop using judgment in
selecting items to examine. The point is that the items selected using judgmental criteria
are not necessarily representative of the population, and conclusions based on items
selected judgmentally should not be extended to the rest of the population.
In establishing the extent of an audit test, an auditor commonly first identifies a stratum of
large or unusual items and plans to examine all of them. If the remaining items are
sampled, this approach usually allows smaller sample sizes and is effectively equivalent to
stratification, which was discussed under Planning and designing the sample .
However, while the large and unusual items are being tested, they are not being sampled,
they are being 100 per cent examined. Audit sampling is being used only for the remaining
items. ASA 500.A56 (ISA 500.A56) explains the point as follows:
While selective examination of specific items from a class of

transactions or account balance will often be an efficient means of
gathering audit evidence, it does not constitute audit sampling.
The results of procedures applied to items selected in this way
cannot be projected to the entire population; accordingly, selective
examination of specific items does not provide audit evidence
concerning the remainder of the population.
QUICK REVIEW
1. The auditor should select sample items so as to have a reasonable
expectation that all sampling units in the population have a chance of
selection.
2. Three selection methods commonly used are:
random selection
systematic selection
haphazard selection.
3. Unacceptable sample selection methods include:
block selection
judgmental selection, where the auditor uses judgmental criteria, such as
large or unusual items, for selection.
4. While judgmental selection will often be an efficient means of gathering
audit evidence, it does not constitute audit sampling.
Page 435
LO 10.6 Performing the audit procedures and evaluating

sampling results
The auditor should apply audit procedures that are appropriate for the particular audit
objective to each item selected. If the application of the procedure is not appropriate for a
selected item, the procedure is ordinarily performed on a replacement item (ASA 530.9–
10/ISA 530.9–10). For example, a cancelled sales invoice may be selected when testing for
evidence of existence of sales, in which case it would be appropriate for the auditor, once
satisfied that the sale had been properly cancelled and did not constitute an error, to
examine a replacement.
Sometimes the auditor is unable to complete a planned audit procedure on a selected item.
If suitable alternative procedures cannot be performed on that item, the auditor usually
considers that item to be a deviation from a control for a test of controls or a misstatement
for substantive tests (ASA 530.11/ISA 530.11). An example of a suitable alternative
procedure might be the examination of subsequent receipts when no reply has been
received in response to a positive confirmation request.
The auditor undertakes an evaluation of the sample results to determine whether the
preliminary assessment of the relevant characteristic of the population ought to be accepted
or rejected. This will be determined on the basis of whether the tolerable deviation rate or
tolerable misstatement has been exceeded. Rejection would lead the auditor to conclude
that, unless further audit evidence is obtained:
the preliminary assessment of control risk cannot be accepted (in the case of tests of
controls), or
the relevant account balance or class of transactions is materially misstated (in the case of
substantive procedures).
The auditor should consider the nature and cause of deviations or misstatements identified
and their possible effect on the particular audit objective and on other areas of the audit
(ASA 530.12–13/ISA 530.12–13). In analysing the errors discovered, the auditor may
observe that many have a common feature; for example, type of transaction, location,
product line or period of time. In such circumstances, the auditor may decide to identify all
items in the population that possess the common feature and extend audit procedures in
that stratum.
For tests of details, misstatements found in the sample should be projected to the
population, a process commonly referred to as projecting misstatements (ASA
530.14/ISA 530.14). This is commonly referred to as extrapolation of sample results. The
auditor estimates the total error for the population in order to obtain a broad view of the
scale of possible errors, and compares this to the tolerable misstatement. The auditor can
project errors in a number of ways. A simple way, and the one that is advocated in the
auditing standards (but which does not take into account sample size or other elements
affecting sampling risk) is to directly project a misstatement rate from a sample to a
population, as illustrated in Figure 10.3 . For example, if the auditor finds a total
misstatement of $10 000 when sampling 20 per cent of a population, this projects to a total
error of $50 000 for 100 per cent of the sample (the population). For tests of controls, no
explicit projection of deviations is necessary since a sample deviation rate expressed as a
percentage (for example, 2 per cent of items deviate from the control condition) is the
projected deviation rate of the population as a whole (ASA 530.A20/ISA 530.A20).
FIGURE 10.3 Projecting misstatements from sample to population
There is one circumstance under which the auditor does not need to project an identified
misstatement or deviation. In accordance with ASA 530.13 (ISA 530.13), under the
extremely rare circumstances that the auditor considers a misstatement or deviation
discovered in a sample to be an anomalous error, the auditor is required to obtain a high
degree of certainty that such misstatement or deviation is not representative of the Page 436
population, and then does not need to project this misstatement or deviation. The
auditor obtains this high degree of certainty by performing additional audit procedures to
obtain sufficient appropriate audit evidence that the misstatement or deviation does not
affect the remainder of the population.
The closer the total projected misstatement is to the auditor’s tolerable misstatement, the
more closely the auditor needs to consider the persuasiveness of the sample results as audit
evidence in light of the objectives of the particular audit procedure and the results of other
audit procedures.
If the evaluation of the sample results indicates that the preliminary assessment of the
relevant characteristic of the population ought to be rejected (total projected misstatements
> tolerable misstatements), the auditor takes one or both of the following actions (ASA
530.A23/ISA 530.A23):
requests that management investigate identified misstatements and the potential for
further misstatements and make any necessary adjustments
modifies planned audit procedures; for example, in the case of a test of controls, the
auditor might test an alternative control or reduce reliance on controls and increase
related substantive procedures.
QUICK REVIEW
1. The auditor applies audit procedures that are appropriate for the particular
objective to each item selected.
2. The auditor considers the nature and cause of any misstatements or
deviations identified.
3. The auditor projects misstatements or deviations found in the sample to
the population, compares this to tolerable misstatements or deviation rate
and considers the effect of the projected error on the particular audit
objective and on other areas of the audit.
LO 10.7 Sampling for tests of controls: in particular,
attribute sampling
In this section we run through the stages of applying sampling to tests of controls. While
Chapter 8 outlines the types of tests of controls related to assertions at risk that the
auditor would undertake, it does not outline how many items to examine, which items to
select, and how sample results are evaluated and extrapolated to the population. A
sampling technique that is suitable for tests of controls and addresses these issues is attrib
ute sampling . This involves examining documents for particular attributes and
providing a given level of confidence on these attributes. Controls for which there is
documentary evidence—such as requirements to match documents, or evidence of
authorisation—are attributes or characteristics of a transaction.
Using attribute sampling, the auditor then examines these characteristics and is able to
estimate the rate of deviations from the prescribed controls in a population, and the results
can be used to support or refute an initial assessment of control risk.
Audit objectives
Use of audit sampling forces the auditor to be specific, before testing, about certain
relationships. For example, the following steps are necessary in considering the
relationship between the sample and the objective of the test:
Identify relevant control objectives, policies and procedures The auditor needs to
specify the control objectives, and the internal control policies and procedures that
achieve them, that are relevant in establishing reliance on the internal controls and
therefore reducing substantive tests of the related account balances. For each transaction
class, the auditor identifies the control objectives that, if achieved, permit an assessed
level of control risk at less than high. For example, for both credit sales and cash Page 437
receipts, relevant control objectives include those related to the occurrence,
completeness and measurement of transaction processing. Controls that may be tested by
attribute sampling are those which are applied to individual transactions, such as
authorisation and the requirement to match documents. Other higher level controls that
are at an organisational level (rather than a transaction level), such as segregation of

duties or codes of ethics, are tested by non-sampling techniques, such as observation and
enquiry.
Identify population and sampling unit In tests of controls, identifying the population
includes specifying the period covered by the test. In tests of credit sales, for example, the
auditor might identify the population as all credit sales transactions for a financial year. A
sampling unit is any of the individual items included in the population. In this example, a
sampling unit would be each individual sales invoice. The auditor also needs to establish
that the population is complete. Usually this is accomplished by footing (adding the total
of all the sampling units). For example, the auditor might establish that the dollar total of
invoices in a sales journal agrees with the dollar total of sales posted to the general
ledger. This step is important because the auditor must establish correspondence between
the sample frame and the population to validly project conclusions from the sample to the
population. Naturally, the appropriate population and sampling unit are influenced by the
objective of the test. For example, if the auditor is concerned with a control objective
related to completeness of sales transactions, the auditor would not sample from the sales
invoices but rather from other forms of evidence, such as shipping documents, to ensure
that a sales invoice has been raised for all goods shipped.
Define the characteristic of interest Tests of controls use a characteristic of interest
that is a condition or an attribute. An attribute either exists or does not exist, which means
that a control activity has either been complied with or not complied with. For example,
if the control is authorisation, the attribute is evidence of the authorisation by a person
who is approved to authorise. In many situations, samples for testing controls are used for
dual-purpose tests (both a test of controls and a substantive test are applied to the same
sample items). For example, the auditor selects a sample of sales invoices, inspects them
to see whether they are initialled to indicate recomputation (test of controls) and re-
performs the computation to check the correct recording of the dollar balance
(substantive test).
Control risk assessment and tolerable deviation

rate
The maximum rate of deviation that would support the auditor’s planned assessed level of
control risk is the tolerable deviation rate. In general, a lower reliance on internal controls
(a higher control risk) will result in an increase in the tolerable deviation rate, and a
decrease in the sample size for tests of controls. In practice, auditors usually assess control
risk across three levels: high, medium and low. For ‘high’, the auditor would omit tests of
controls.
The next decision the auditor has to make is what tolerable deviation rates to associate with
the levels of control risk. Auditors sometimes use tolerable deviation rates of 0 to 5 per
cent for a low level of control risk (high reliance on the control), 6 to 10 per cent for a
medium level of control risk, and assess control risk at the high level when the rate exceeds
10 per cent (no reliance on the control).
Such deviation rates may at first seem relatively high. However, it must be remembered
that a deviation does not necessarily mean that a monetary misstatement has resulted.
Table 10.1 relates tolerable deviation rates and planned assessed level of control risk to
the risk model outlined in Chapter 4 . Control risk is expressed as a specific percentage
only when statistical sampling is being used. However, the quantitative model makes the
relationships easier to visualise.
RELATIONSHIP BETWEEN TOLERABLE DEVIATION
TABLE 10.1
RATE AND PLANNED LEVEL OF CONTROL RISK
Qualitative level of control risk Tolerable deviation rate (%)
Low 0 to 5
Medium 6 to 10
High Omit test
Page 438
Determining sample size

As outlined in ASA 530.A11 (ISA 530.A11), sample size can be determined by the
application of a statistically based formula or through the exercise of professional judgment
objectively applied to the circumstances.
Using a statistically based formula allows the computation of a maximum rate of deviation
at a specified confidence level, given a rate of occurrence in a sample of size n. In tests of
controls, an occurrence is a deviation from prescribed policies or procedures. Rather than
introduce the traditional symbols for the statistical model and then equate them with the
audit sampling terms, the formula for the maximum deviation rate is presented using audit
sampling terms and simple letter symbols, as follows:
R
Maximum deviation rate (MDR) =
n
where R = reliability factor for desired level of assurance

n = sample size.
The reliability factor (R) that corresponds to a desired level of assurance for an expected
number of deviations is found in Table 10.2 . For example, the reliability factor for a
desired level of assurance of 95 per cent with no expected deviations is 3.00.
RELIABILITY FACTORS FOR ASSESSING DESIRED
TABLE 10.2
LEVEL OF ASSURANCE
Number of errors Desired level of assurance
99% 95% 90%
0 4.61 3.00 2.31
1 6.64 4.75 3.89
2 8.41 6.30 5.33
3 10.05 7.76 6.69
4 11.61 9.16 8.00
5 13.11 10.52 9.28
Source: Adapted from Australian Accounting Research Foundation (AARF), 1983, p. 99.
The related formula for minimum sample size to achieve a given tolerable deviation rate is
as follows:
R
sample size (n) =
T DR
where TDR = tolerable deviation rate.
To use the model to estimate sample size in planning tests of controls, the auditor needs to
specify the following:
the tolerable deviation rate

the desired level of assurance
the expected number of deviations.
Use of the statistical model to estimate sample size can best be explained by an example
(see Global example 10.3 ).
Page 439
GLOBAL EXAMPLE 10.3 Determining sample size for attribute
sampling
Assume an auditor has specified a tolerable deviation rate of 7 per cent

(0.07) and a desired level of assurance of 90 per cent, and expects to find at
most one deviation. Using the formula, a minimum sample size of 56 could
be computed as follows:
R Reliability factor
n = =
T DR Tolerable deviation rate
3.89
=
0.07
= 56
Notice that 3.89 is the reliability factor for one expected deviation at a 90 per
cent desired level of assurance, from Table 10.2 .
Sample size can also be determined using Tables 10.3 and 10.4 , which assess the
expected deviation rate in percentage terms. Assume the auditor expects to find at most a 1
per cent deviation rate in the population. By referring to Table 10.3 , using a tolerable
deviation rate of 7 per cent (as in Global example 10.3 ) and a desired level of
assurance of 90 per cent, a minimum sample size of 56 (the 55 in Table 10.3 is due to
rounding) is required (as calculated above) and one deviation (the figure in parentheses) is
expected.
STATISTICAL SAMPLE SIZES FOR 90 PER CENT DESIRE
TABLE 10.3
OF ASSURANCE
Number of expected deviations/misstatements in parenthesis
Expected Tolerable deviation/misstatement rate

deviation/
misstatement 2% 3% 4% 5% 6% 7% 8% 9%
rate (%)
0.00 114(0) 76(0) 57(0) 45(0) 38(0) 32(0) 28(0) 25(0
0.25 194(1) 129(1) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1
0.50 194(1) 129(1) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1
0.75 265(2) 129(1) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1
1.00 * 176(2) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1
1.25 * 221(3) 132(2) 77(1) 64(1) 55(1) 48(1) 42(1
1.50 * * 132(2) 105(2) 64(1) 55(1) 48(1) 42(1
1.75 * * 166(3) 105(2) 88(2) 55(1) 48(1) 42(1
2.00 * * 198(4) 132(3) 88(2) 75(2) 48(1) 42(1
2.25 * * * 132(3) 88(2) 75(2) 65(2) 42(1
2.50 * * * 158(4) 110(3) 75(2) 65(2) 58(2
2.75 * * * 209(6) 132(4) 94(3) 65(2) 58(2
3.00 * * * * 132(4) 94(3) 65(2) 58(2
3.25 * * * * 153(5) 113(4) 82(3) 58(2
3.50 * * * * 194(7) 113(4) 82(3) 73(3

3.75 * * * * * 113(4) 98(4) 73(3
4.00 * * * * * 149(6) 98(4) 73(3
5.00 * * * * * * 160(8) 115(6
6.00 * * * * * * * 182(1
7.00 * * * * * * * *
* Sample size is too large to be cost effective for most audit applications.
Note: This table assumes a large population (> 500 items).
Source: Adapted from AARF, 1983, p. 91.
Page 440
STATISTICAL SAMPLE SIZES FOR 95 PER CENT DESIRE
TABLE 10.4
LEVEL OF ASSURANCE
Number of expected deviations/misstatements in parenthesis
Expected Tolerable deviation/misstatement rate

deviation
/misstatement 2% 3% 4% 5% 6% 7% 8% 9%
rate (%)
0.00 149(0) 99(0) 74(0) 59(0) 49(0) 42(0) 36(0) 32(0)
0.25 236(1) 157(1) 117(1) 93(1) 78(1) 66(1) 58(1) 51(1)
0.50 * 157(1) 117(1) 93(1) 78(1) 66(1) 58(1) 51(1)
0.75 * 208(2) 117(1) 93(1) 78(1) 66(1) 58(1) 51(1)
1.00 * * 156(2) 93(1) 78(1) 66(1) 58(1) 51(1)
1.25 * * 156(2) 124(2) 78(1) 66(1) 58(1) 51(1)
1.50 * * 193(3) 124(2) 103(2) 66(1) 58(1) 51(1)
1.75 * * 227(4) 153(3) 103(2) 88(2) 77(2) 51(1)
2.00 * * * 181(4) 127(3) 88(2) 77(2) 51(1)
2.25 * * * 208(5) 127(3) 88(2) 77(2) 51(1)
2.50 * * * * 150(4) 109(3) 77(2) 68(2)
2.75 * * * * 173(5) 109(3) 95(3) 68(2)
3.00 * * * * 195(6) 129(4) 95(3) 84(3)
3.25 * * * * * 148(5) 112(4) 84(3)
3.50 * * * * * 167(6) 112(4) 84(3)

3.75 * * * * * 185(7) 129(5) 100(4)
4.00 * * * * * * 146(6) 100(4)
5.00 * * * * * * * 158(8)
6.00 * * * * * * * *
7.00 * * * * * * * *
* Sample size is too large to be cost effective for most audit applications.
Note: This table assumes a large population (> 500 items).
Source: Adapted from AARF, 1983, p. 90.
Table 10.3 provides sample sizes where the desired level of assurance is 90 per cent
(which, as outlined in Table 10.1 , can equate to an initial assessment of control risk as
medium), while Table 10.4 provides sample sizes where the desired level of assurance
is 95 per cent (which, as outlined in Table 10.1 , can equate to an initial assessment of
control risk as low).
Selection of sample items for tests of controls

Having determined the appropriate sample size, it is then a matter of determining which
sample items to select. The representative selection methods of random selection and
systematic selection discussed earlier generally apply to both tests of controls and
substantive tests. However, stratification is not usually applicable to tests of controls, as it
is commonly based on the recorded amount of items included in an account balance.
Usually, the same control and processing activities apply to transactions in a class
regardless of the dollar amount of the transactions. If a particular type of transaction in a
class is subject to different processing or different control activities, tests of these Page 441
transaction types should be planned and evaluated separately. For example, if a

company makes both retail and wholesale sales, and retail sales are subject to different
processing steps and control activities, samples for testing controls associated with retail
sales and wholesale sales should be designed separately.
Systematic selection (defined earlier in this chapter) is often useful for tests of controls
because it helps to achieve the auditor’s internal control objective of testing continuity of
controls by ensuring sampling is continuous throughout the year.
Evaluation of sample results

The auditor applies audit procedures to the selected items and determines the deviation rate
of the sample. If the auditor cannot apply the planned procedure (or an effective
alternative) to the item, it should be counted as a deviation in evaluating the sample. The
auditor then estimates the population deviation rate to obtain a broad view of the scale of
possible deviations and compares this to the tolerable deviation rate.
The sample deviation rate is the auditor’s best estimate of the population deviation rate. For
example, the auditor selects a sample of 25 items, applies the planned audit procedure and
finds one deviation. The sample deviation rate is therefore 4 per cent (1 ÷ 25), and the best
estimate of the population deviation rate is also 4 per cent. If the tolerable deviation rate
was established as, say, 5 per cent (anything greater than 4 per cent) for the planned
assessed level of control risk, the sample result supports the auditors’ reliance on the
internal control. This simple approach is currently advocated in the auditing standards
(ASA 530.A20/ISA 530.A20), but may be criticised as it does not take into account
sampling risk (or the auditor’s desired level of assurance). This is even more of a concern
with the small sample sizes that are commonly used in practice. For example, if the auditor
had sampled 20 transactions and found one deviation (5 per cent sample deviation rate),
rather than 60 transactions with three deviations found (still 5 per cent sample deviation
rate), the risk that the sample was not representative of the population (sampling risk)
would have been higher with the smaller sample size.
The auditor’s evaluation of sample results is not limited to quantitative analysis. A

projection (commonly termed extrapolation) of sample results and consideration of
sampling risk are required for all audit samples (quantitative aspects), but the auditor also
needs to consider the nature and cause of the deviations. When analysing deviations, the
auditor may observe that many have a common feature; for example, type of transaction,
location, product line or period. In such circumstances the auditor may decide to identify
all items in the population that possess the common feature and extend audit procedures in
that stratum (ASA 530.A17/ISA 530.A17).
QUICK REVIEW
1. The types of tests of controls for which audit sampling is especially useful
are those which are associated with individual transactions and involve the
inspection or matching of source documents.
2. In determining sample size, the auditor can either use professional
judgment or the application of a statistically based formula. If using a
statistically based formula, the auditor needs to specify the tolerable
deviation rate, desired level of assurance and expected number of
deviations.
3. Samples for tests of controls are commonly selected using systematic
selection techniques, which help ensure that controls are tested
throughout the entire period.
4. The sample deviation rate is the auditor’s best estimate of the population
deviation rate. If the sample deviation rate is greater than the tolerable
deviation rate, the auditor cannot rely on the internal control being
assessed. It should be recognised that this method does not explicitly
consider sampling risk.
Page 442
LO 10.8 Sampling for substantive tests: in particular,

dollar-unit sampling
Audit sampling as it applies to substantive tests has the objective of determining the dollar
amount of misstatement in an account. While Chapter 9 outlines the types of
substantive tests related to assertions at risk that an auditor would undertake, it does not
outline how many items the auditor should examine, which items to select, and how the
auditor extends a conclusion reached from testing a sample of the items to the balance of
the account. These issues are addressed under this learning objective.
The sampling approach for substantive tests that weights individual transactions or account
balances evenly is known as the traditional sampling approach. Another approach that is
commonly used in practice is the dollar-unit sampling approach (DUS), which is a form of
monetary unit sampling where the monetary units are dollars. The difference between
traditional sampling and DUS is in the definition of the sampling unit. Under traditional
sampling, the sampling unit is, for example, each inventory item on the inventory or each
debtor contained in the accounts receivable master file.
Thus, rather than giving each debtor an equal chance of being selected, the DUS approach
gives each dollar an equal chance of being selected. Under traditional sampling, debtor A
and debtor B have an equal chance of selection, whereas under DUS, if debtor A owes
$1000 and debtor B owes $100, debtor A is 10 times more likely to be selected. Thus DUS
directs the auditor’s attention to the more significant accounts or line items. As the auditor
then tests the debtor (for example, through confirmation) or inventory item (for example,
through sighting) that the selected dollar belongs to, rather than the individual dollar, it
increases the auditor’s dollar coverage, which is an important advantage when undertaking
substantive tests.
DUS is beneficial in testing specific assertions, such as existence, which are related to
overstatement. Under DUS, an overstatement will increase the likelihood of an account

being selected. Thus, if the true value of a debtor’s balance is $10 000, but it is overstated
in the accounts at $50 000, this overstatement will increase the likelihood of the account
being selected. However, DUS will direct the auditor’s attention away from
understatements, related to the completeness assertion. If the true debtor’s balance of $10
000 is understated so that it is in the accounts at $2000, the number of sampling units has
been decreased. Thus, DUS is seen as an appropriate sampling approach for testing the
existence assertion, but another testing procedure would need to be taken for testing the
completeness assertion.
Planning for substantive tests

The following matters should be considered in planning the sample:
the relationship of the sample to the relevant audit objective

preliminary judgments about materiality levels (tolerable misstatement)
the auditor’s desired level of assurance that actual misstatement is not greater than
tolerable misstatement
characteristics of the population, that is, the items comprising the account balance
the use of other substantive procedures directed at the same financial report assertion.
The relationship of the sample to the relevant audit

objective
The objectives of undertaking substantive tests should be related to the specific audit
assertion that is being addressed. Because of the characteristics of misstatements that can
be detected by sampling, the relevant assertions are most commonly those related to
existence or accuracy, valuation and allocation. For example, by selecting from a list of
items included in an account balance and applying audit procedures to them, the auditor
could detect fictitious items or items incorrectly included in the balance (existence), or
incorrectly valued items. However, the auditor could not detect items that should be there
but are omitted from the balance (completeness).
Generally, the other decisions in this step are similar to those for the comparable Page 443
step in planning tests of controls. The auditor should identify the population and
sampling unit; specify the population from which sample items will be selected; and
establish that the population is complete (usually by footing (adding) the list or master file).
Naturally, the characteristic of interest is a monetary misstatement of the account balance.
An aspect of identifying the population that is unique to substantive tests is the undertaking
of a preliminary analysis of the account balance to identify individually significant items.
An item could be significant because of its size or its nature. Essentially, anything that
causes the auditor to believe that particular items in the account balance are subject to a
greater risk of material misstatement could cause those items to be considered significant.
For example, an analysis of sales returns related to defective goods by product and location
may indicate that certain inventory items are more likely to be misstated. The portion of
the account balance remaining after stratifying individually significant items and
undertaking 100 per cent testing on these items then becomes the population subject to
sampling.
Preliminary judgments about materiality levels (tolerable

misstatement)
The tolerable misstatement for an account balance is, in effect, the materiality assessment
at the account balance level. It is the maximum amount of monetary misstatement that can
exist in an account balance that, when combined with monetary misstatement in other
account balances, would not cause the financial report taken as a whole to be materially
misstated. As tolerable misstatement increases, the necessary sample size decreases.
Desired level of assurance
The desired level of assurance is the level of assurance the auditor desires that the results of
the sample are in fact indicative of the actual incidence of misstatement in the population.
In planning the sample for a substantive test of balances, the desired level of assurance is
related to the detection risk for the test. The auditor assesses inherent risk and control risk
and establishes detection risk at the level appropriate to keep the audit risk for the account
balance at an appropriately low level. Notice that it is audit risk and not detection risk that
needs to be held to this appropriately low level. If the auditor concludes that the likelihood
of material misstatement in the account balance is low, the detection risk for the direct test
of a balance, using audit sampling, can be relatively high. As the desired level of assurance
increases, the necessary sample size increases.
There is another aspect of sampling risk for substantive tests—the risk of incorrect
rejection. It is possible that the sample supports the conclusion that the account balance is
materially misstated when it is not. This aspect of sampling risk is not commonly specified
as a consideration in planning samples because it relates to audit efficiency rather than
effectiveness. An auditor who incorrectly concludes that an account balance is materially
misstated usually expands substantive tests and ultimately reaches the appropriate
conclusion.
Population characteristics: expected misstatement,
variability and population size
The characteristics of the population that are relevant to a substantive test of balances are
expected misstatement, variability and population size. Generally, as the size and frequency
of expected misstatement increases, or as the variability or population size increase, larger
sample sizes are necessary. However, expected misstatement and variability have a much
more significant influence on sample size than the size of the population. The auditor can
reduce the effects of variability in the population by stratifying the population.
The use of other substantive procedures directed at Page 444
the same financial report assertion
The more the auditor relies on other substantive procedures (either tests of detail or
analytical procedures) to reduce to an acceptable level the detection risk regarding a
particular account balance or class of transactions, the less assurance is required from
sampling and, therefore, the smaller the sample size can be.
Determination of sample size for substantive tests

As was outlined for tests of controls, sample size can be determined by the application of a
statistically based formula or through the exercise of professional judgment or audit firm
decision aids objectively applied to the circumstances.
If using a statistically based formula, recall that the underlying statistical model estimates
sample size by dividing the appropriate reliability factor by the tolerable misstatement
(TM). The same method can be used for a substantive test of a balance by converting
tolerable misstatement to a rate. If the population subject to sampling is designated as book
value (BV), then the ratio of tolerable misstatement to book value is the tolerable
misstatement rate. The formula presented earlier then becomes:
R Reliability factor
n = =
T M ÷ BV Tolerable misstatement ÷ Book value
For convenience, this is usually presented as:
BV × R
n =
TM
Assume that the portion of the balance subject to sampling is $1 000 000 (BV), the
tolerable misstatement is $50 000 (TM), there is zero expected misstatement, and the
desired level of assurance is 95 per cent (which equates to a risk of incorrect acceptance of
5 per cent). An appropriate reliability factor (R) of 3.00 can be identified from
Table 10.2 . The sample size (n) can then be estimated as follows:
$1 000 000 × 3.00
n = = 60
$50 000
Selecting the reliability factor
The reliability factors that are used are the same for both tests of controls and substantive
tests. The relationship between reliability factors and specific percentages of the risk of
incorrect acceptance is shown in Table 10.5 (this is a variation of Table 10.2 for
zero expected misstatements).
RELATIONSHIP BETWEEN RELIABILITY FACTOR

TABLE 10.5
AND RISK OF INCORRECT ACCEPTANCE
Risk of incorrect acceptance (%) Reliability factor
1.0 4.6
5.0 3.0
7.5 2.6
10.0 2.3
The reliability factors outlined in Table 10.5 for the specified risk levels are for zero
expected misstatements. However, the auditor is required to consider the level of expected
misstatement in the population (ASA 530 Appendix 3/ISA 530 Appendix 3). The impact of
the relaxation of this assumption on sample size is shown in Tables 10.3 and 10.4 .
Page 445
Selection of sample items for substantive tests
The representative selection methods of random, systematic and haphazard selection are all
appropriate for selecting samples items for substantive tests. However, as mentioned
earlier, some form of stratification may increase efficiency. Stratification reduces the
variability of items within each stratum and therefore allows sample size to be reduced
without a proportional increase in sampling risk. When performing substantive procedures,
an account balance or class of transactions will often be stratified by dollar value, allowing
greater audit effort to be directed to the large-value items, which contain the greatest
potential monetary misstatement.
Using the DUS approach, the sampling units are individual dollar units rather than physical
units. A population of $1 000 000 that contains 1000 accounts is viewed as a population
with 1 000 000 sampling units—the individual dollars. When an individual dollar is
selected, however, it is naturally attached to the other dollars associated with the account
(for example, individual debtor or inventory item), and the selected dollar drags the entire
dollar amount of the account with it. In other words, the dollar unit identifies which
account the audit procedure is used to test.
In practice, the most commonly applied sample selection technique is systematic sampling.
This method of sample selection was outlined earlier in this chapter. To use systematic
selection the auditor needs to establish the cumulative dollar total of items in the
population. A random start is established within the first sampling interval to determine the
first item selected, and the dollar interval is added to determine the next item. The random
start plus two times the sampling interval determines the third item, and so on through the
population. Global example 10.4 demonstrates the use of this DUS selection method.
GLOBAL EXAMPLE 10.4 Illustration of DUS selection using
systematic sampling with a dollar interval
Invoice no. Amount of item Cumulative total
A1001 $ 6 000 $ 6 000
A1002 1 000 7 000
A1003 15 000 22 000
A1004 5 000 27 000
A1005 12 000 39 000
A1006 10 000 49 000
A1007 9 000 58 000
A1008 17 000 75 000
A1009 11 000 86 000
A1010 8 000 94 000
A1011 7 000 101 000
• • •
• • •
• • •
Population size (BV) $1 000 000
Sample size (n) 60
Sampling interval (SI) $ 16 667

($1 000 000 ÷ 60)
Random start (RS) $ 14 068
Items selected $ Invoice no. Amount $
RS 14 068 A1003 15 000
RS + SI 30 735 A1005 12 000
RS + 2SI 47 402 A1006 10 000
RS + 3SI 64 069 A1008 17 000
RS + 4SI 80 736 A1009 11 000
RS + 5SI 97 403 A1011 7 000
• • • •
• • • •
• • • •
The invoice numbers and dollar amounts of the items are listed in the client’s
accounting records. The cumulative total is computed by the auditor.
Assume the random start is the 14 068th dollar. This dollar is contained in
item A1003, so that is the first item selected and the audit procedure is
applied to this invoice. The calculated sampling interval (16 667) is added to
the random start to determine the next item to be selected. The 30 735th
dollar is contained in item A1005, so that is the next item selected. The
selection process continues in this manner through the population. Notice
that all items larger than the sampling interval will always be selected. This
process is usually automated using audit software (described in
Chapter 9 ).
Page 446
Evaluation of sample results for substantive tests

After the planned audit procedures have been applied to selected items, the auditor may
have detected a level of monetary misstatement. This misstatement has to be projected to
the population from which the sample was selected.
There are simple and complex approaches to evaluating sample results. The guidance in the
auditing standards suggests that, for substantive testing, an unexpectedly high level of
monetary misstatement in a sample may cause the auditor to believe that an account
balance or class of transactions is materially misstated (ASA 530.A22/ISA 530.A22). The
standard requires the auditor to project monetary misstatements found in a sample to a
population. This projection is then compared to the tolerable misstatement that has been set
by the auditor for that population. There are several ways to do this. One way is outlined in
Global example 10.5 .
Page 447
GLOBAL EXAMPLE 10.5 Evaluation of sample results for
substantive test
Assume that an auditor has examined the supporting documentation for

additions to property, plant and equipment, and found a net monetary
overstatement in the sample. The relevant information is as follows:
Amount of additions $1 000 000
Number of additions 1 000
Sample size 60
Dollar value of sample $100 000
Tolerable misstatement $50 000
Misstatement found in sample $2 000
For simplicity, it is assumed that the auditor did not identify individually
significant items or stratify the items, so that the total dollar amount of
additions and the population subject to sampling are the same. Using dollar-
unit sampling, where each dollar is weighted evenly, the projected
misstatement is $20 000 ($2000 × 1 000 000 ÷ 100 000).
The auditor next compares projected misstatement to tolerable

misstatement and considers the sampling risk. In the example given,
projected misstatement of $20 000 is less than tolerable misstatement of
$50 000, and the auditor would conclude that the additions to plant and
equipment are not materially misstated.
The method of evaluating sample results outlined in Global example 10.5 is in

accordance with auditing standards (ASA 530.A18–A20/ISA 530.A18–A20) and is widely

used in practice. However, it does not take account of sampling risk, the risk that this result
might be obtained from a sample of $100 000 from a population of $1 000 000, although
the true monetary misstatement may exceed tolerable misstatement (except to the extent
that the auditor is asked to consider sampling risk after undertaking this calculation). This
is even more of a concern with the small sample sizes that are currently widely used in
practice. For example, if we had sampled $50 000 out of $1 000 000, rather than $100 000,
as above, there would have been an even greater possibility that the sample did not
accurately reflect the population.
The auditor also needs to consider the qualitative aspects of sample error: the nature and
cause of the misstatement and the implications for other phases of the audit. This analysis
may allow the auditor to focus on a particular cause of misstatement. For example, if the
auditor finds that all misstatements in additions to property, plant and equipment occurred
in only one division out of four in the company, the auditor would likely extend testing only
in the particular division where misstatements were identified.
Finally, the auditor needs to relate the sample evaluation to the results of other audit
procedures relevant to the account balance and combine the results of all audit tests. The
auditor then evaluates the effect on the financial report taken as a whole.
If the auditor initially concludes that an account balance is materially misstated, a common
approach is to expand the portion of the account balance that is 100 per cent examined. The
auditor identifies individually significant items and examines all of them. The remaining
population is sampled. Any monetary misstatement in items examined 100 per cent is
known misstatement, and the known misstatement can readily be corrected by the client.
In DUS using systematic selection, all items larger than the sampling interval will by
definition be included in the sample (as they must contain a sampling unit) and be
examined. Monetary misstatement in these items is known misstatement and is not
projected. It is also possible that items larger than the sampling interval may be selected
more than once if systematic sampling is used, or any items with more than one dollar may
be selected more than once using random numbers. Thus, by testing the one transaction or
account balance, it is possible that more than one sampling item can be tested. For this
reason the actual sample size, in terms of transactions or account balances, may be smaller
than the planned sample size.
QUICK REVIEW
1. Sampling for substantive tests is usually aimed at the assertions of
existence and accuracy, valuation and allocation.
2. The characteristic of interest is monetary misstatement of the account
balance.
3. Dollar-unit sampling directs the auditor’s attention towards large-value
items and items that are overstated, but away from items that are
understated.
4. Systematic sampling with a dollar interval is the most commonly Page 448
applied method for sample selection when undertaking DUS.
5. Stratification is widely used to allow greater audit effort to be directed to
the large-value items.
6. The auditor is required to project the misstatements found in the sample to
the population. The auditor concludes that an account balance is not
materially misstated if projected total misstatement is less than tolerable
misstatement.
LO 10.9 Other statistical sampling approaches
The other statistical models that can be used in substantive tests differ in many respects
from the DUS model. Generally, these models are referred to as classical statistical
sampling methods because they all use the normal distribution and related mathematics.
They are also called variables sampling methods, because the characteristic of interest
is a variable rather than an attribute, and estimation sampling methods, because they
are used to estimate the characteristic that is of interest. In auditing, the characteristic of
interest is monetary misstatement, and the estimation sampling methods are used in a
hypothesis-testing format—does monetary misstatement in a recorded balance exceed a
material amount? Further details on estimation sampling using an audit-hypothesis-testing
approach are given on the website associated with this book.
The auditor uses sampling to create an estimate of the amount, and compares it to the
amount recorded by the client. If the estimate is reasonably close to the recorded amount,
the sampling procedure has provided audit evidence supporting the balance. The primary
classical statistical sampling methods used in auditing are mean-per-unit estimation,
difference estimation and ratio estimation.
Mean-per-unit estimation
In the mean-per-unit estimation technique, the auditor selects sample items from the
population and determines the audited value of each item selected. The auditor calculates
the average value (the mean) of the items and multiplies the average value by the number
of items in the population. This type of estimate is called a simple extension. If the average
value of the items is $1375 and there are 1000 items in the population, the estimated value
is $1 375 000. The mean times the number of items is called the point estimate. The audit
sample assures the auditor that the point estimate plus or minus the precision limit at a
specified reliability percentage includes the true value of the population. For example,
using this technique the auditor can say, ‘I am 95 per cent confident that the range of $1
375 000 plus or minus $38 450 (the precision limit, which would be determined by
reference to the standard deviation of the sample) includes the actual value of the
inventory’. Mean-per-unit estimation can be applied to individual strata or to populations.
Difference estimation
In difference estimation , the auditor calculates the difference between the audited
value and the book value of each item in the sample. The average of the differences is
calculated and then multiplied by the number of items in the population to obtain an
estimated total projected difference. The population value is then estimated as the total
book value plus or minus the total projected difference. The sample sizes when using this
method tend to be much smaller than those where mean-per-unit estimation is being used.
Ratio estimation
Ratio estimation is similar to difference estimation. For each item selected, the auditor
determines an audited value and a book value. In ratio estimation, the sum of all of the
book values of the sampled items is calculated, and the sum of all of the audited Page 449
values of the sampled items is calculated. The auditor calculates a ratio by
dividing the sum of the audited values by the sum of the sample book values. The ratio is
multiplied by the recorded total book value for the account balance to create the estimated
audited value.
As in difference estimation, achieved precision is calculated (using the standard deviation

of the individual ratios in the sample). The auditor calculates a range—the estimated
audited value plus or minus the achieved precision—and determines whether the recorded
book value falls within the range.
QUICK REVIEW
1. Other sampling approaches are also used in substantive testing. These
are commonly called variables sampling methods, because the
characteristic of interest is a variable rather than an attribute, and
estimation sampling methods, because they are used to estimate the
characteristic that is of interest—the amount of misstatement.
2. The primary classical statistical sampling methods used in auditing are:
mean-per-unit estimation
difference estimation
ratio estimation.
Summary
Even in today’s world of advanced data analytics and big data, audit sampling is a very
important part of the evidence-gathering procedures of the audit process. It is through
sampling techniques that the auditor can determine the number of items in a population
that need to be tested and which particular items in the population will be selected for audit
testing. Further, after audit testing has been carried out, the auditor can extrapolate sample
results so that a conclusion can be formed on the population.
Key terms
attribute sampling
audit sampling
block selection
characteristic of interest
desired level of assurance
difference estimation
dollar-unit sampling (DUS)
estimation sampling
expected misstatement
expected rate of deviation
haphazard selection
judgmental selection based on sample item characteristics
mean-per-unit estimation
non-sampling risk
non-statistical sampling
population
projecting misstatements
random selection
ratio estimation
reliability factor
sampling risk
sampling unit
statistical sampling
stratification
systematic selection
tolerable misstatement
tolerable rate of deviation
variables sampling
Page 450
References and additional readings

American Institute of CPAs (AICPA) (2012) Audit Guide: Audit Sampling, New York.
Australian Accounting Research Foundation (AARF) (1983) Audit Sampling, Audit Guide
No. 1, Melbourne.
Dworin, L.D. and Grimlund, R.A. (1984) ‘Dollar unit sampling for accounts receivable and
inventory’, The Accounting Review, April, pp. 218–41.
EY (2015) ‘How big data and analytics are transforming the audit’, www.ey.com/gl/en/ser
vices/assurance/ey-reporting-issue-9-how-big-data-and-analytics-are-transforming
-the-audit#item, accessed 29 November 2017.
Hall, T.W., Herron, T.L. and Pierce, B.J. (2006) ‘How reliable is haphazard sampling?’ The
CPA Journal, January, pp. 26–7.
Leslie, D.A., Teitlebaum, A.D. and Anderson, R.J. (1979) Dollar Unit Sampling: A
Practical Guide for Auditors, Copp, Clark & Pitman, Toronto.
Robertson, J.C. and Rouse, R. (1994) ‘Substantive audit sampling—the challenge of
achieving efficiency along with effectiveness’, Accounting Horizons, March, pp. 35–44.
Stewart, M. and Dunn, J. (2000) ‘Introducing audit sampling’, The Auditor’s Report,
Summer, pp. 15–16.
Review questions
Definition and features

10.1 Explain the difference between sampling risk and non-sampling risk. LO 10.1
Various means of gathering audit evidence

10.2 Explain the differences between non-statistical and statistical sampling.
LO 10.2
10.3 Selecting from a list of accounts receivable, the auditor uses their judgment
to bias their selection towards the large-value items. Explain why this is or is
not an example of audit sampling. LO 10.2
10.4 Explain how the concept of big data and advanced data analytics may affect
the auditor’s approach to sampling. LO 10.2
Planning and designing the sample

10.5 Explain the meaning of stratification and why auditors stratify audit
populations. LO 10.3
10.6 Explain whether sampling is equally appropriate as an evidence-gathering
technique for both the assertions of existence and completeness.
LO 10.3

10.7 Identify three factors that influence sample sizes for tests of controls.
LO 10.4
10.8 Explain the relationships between risk, precision, expected error in
population and sample size. LO 10.4
Selecting the sample

10.9 Identify the three requirements for selecting samples under ASA 530.
LO 10.5
10.10 Explain what is meant by random sampling. LO 10.5
Performing the audit procedures and evaluating

sample results
10.11 If the evaluation of sample results suggests that there is material
misstatement in the population, what courses of action are available to the
auditor? LO 10.6
Sampling for tests of controls: in particular,

attribute sampling
10.12 Provide a brief explanation of attribute sampling and why it is used for tests
of controls. LO 10.7
Page 451
Sampling for substantive tests: in particular, dollar-

unit sampling
10.13 Identify and explain for which account balances it is most beneficial to use
dollar-unit sampling. LO 10.8
Other statistical sampling approaches

10.14 What is mean-per-unit estimation? LO 10.9
Discussion problems and case studies
Definition and features

10.15 EASY You are an audit manager and your new audit graduate has asked
you why he is supposed to use sampling, rather than testing all the
transactions.
REQUIRED
Explain to the new graduate the objective of the auditor when using
sampling, and why auditors traditionally sample, rather than test every
transaction. LO 10.1
Various means of gathering audit evidence

10.16 EASY In auditing a very large manufacturing company, for which of the
following audit procedures could you undertake sampling? For those audit
procedures where sampling is not appropriate, explain why.
(a) Checking the average cost prices of inventory items per the stock
records with the year-end inventory schedules
(b) Physically inspecting items of plant and equipment
(c) Checking individual wage records against award rates

(d) Scrutinising the list of payments at the beginning of the next financial
year, looking for expenses possibly incurred, but not accrued in the
accounting records, before the end of the financial year
(e) Examining the documentation that records the docket numbers of
the last goods received
(f) Comparing a supplier’s monthly statement to the balance for that
supplier in the accounts payable ledger
(g) Scrutinising credit notes given to customers after year end LO 10.2
Source: This question was adapted from the Chartered Accountants Program of the Institute of
Chartered Accountants in Australia, 2011 (2) audit and assurance module.
10.17 MEDIUM You are the manager on the audit of Rising Wave Ltd for the
year ended 30 June 2018. Your audit senior, Sally Wakes, has analysed
additions to property, plant and equipment for the year as follows.
NUMBER OF RANGE AMOUNT ($)
ADDITIONS
10 Above $10 000 160 000
35 $1000 to $9999 140 000
45 300 000
Sally plans to vouch all the additions above $10 000 to supporting
documentation and rely on analytical procedures for the remaining amount.
REQUIRED
(a) Is this an acceptable audit approach? Page 452
(b) Explain whether there is any detection risk associated with this
approach.
(c) Explain whether there is any sampling risk associated with this
approach. LO 10.2
Planning and designing the sample

10.18 EASY You are the audit senior on the audit of Alpha Engines Ltd (AEL) for
the year ended 30 June 2018. The auditor’s report needs to be signed on
31 August 2018, and the audit manager has requested you to confirm the
accounts receivable balance. The accounts receivable balance was $15 178
543 as at 30 June 2018 and is considered material.
AEL’s customer payment terms are 30 days from the date of the invoice. 10
customers represent 70 per cent of the accounts receivable balance
outstanding as at 30 June 2018. The remaining 30 per cent is made up of
numerous smaller amounts, most of which have been outstanding for more
than 60 days.
REQUIRED
Identify two key factors that you would consider when designing a
sampling approach for accounts receivable confirmations for AEL’s
accounts receivable balance. LO 10.3
10.19 EASY Your audit client, Speakers & Woofers Ltd, is a retailer of audio
equipment. It carries approximately 300 stock lines, with individual items
ranging in value from $50 to $10 000. Approximately 25 per cent of year-
end inventory is expected to be held at the company’s central warehouse
in Melbourne, with the remainder spread relatively evenly across eight
retail outlets. A full stocktake will be carried out at all locations at year end.
REQUIRED
(a) In order to gather evidence about the existence and completeness
assertions, the auditor will perform test counts at the stocktake. In
general, from which samples should populations be selected to
satisfy each of these two assertions?
(b) Explain which of the two assertions you would pay more attention
to. LO 10.3

10.20 EASY Kaela Stevens is auditing Turbo Ltd, an existing client, for the year
ended 30 June 2018. Turbo’s accounts receivable was recorded at $4 000
000 and comprised more than 1000 customer accounts. However, Turbo’s
10 largest customer balances comprised more than 25 per cent of the
recorded accounts receivable. As a result, Kaela is considering the use of
REQUIRED
Outline four factors Kaela should consider in determining the sample size
for accounts receivable, and indicate how she would determine the value
to place on these factors. LO 10.4
10.21 MEDIUM Ross Warner is auditing the accounts receivable of Blue Ribbon
Ltd for the year ended 30 June 2018. Blue Ribbon’s accounts receivable
balance was recorded at $5 000 000 and comprised more than 2000
customer accounts. However, Blue Ribbon’s 10 largest customer balances
comprised a high percentage of the recorded accounts receivable (more
than $1 500 000, or 30 per cent). As a result, Ross is considering using
Based on prior audits and other judgments, Ross has established the
following parameters:
Risk of incorrect acceptance 5%
Tolerable error $300 000
Expected error $100 000
REQUIRED Page 453

Calculate the necessary sample size and sampling interval that would be
used by Ross in the audit of Blue Ribbon. LO 10.4
Selecting the sample

10.22 EASY You are the audit manager on the audit of Whirlpool Ltd for the
year ended 30 June 2018 and your new audit graduate, Otto Gray, has
selected the four weeks in May to test for payroll.
REQUIRED
Explain to Otto why the sample selection method adopted is not
appropriate. LO 10.5
Performing the audit procedures and evaluating

sample results
10.23 MEDIUM You are reviewing the audit working papers prepared by a
junior auditor for the audit for the year ended 30 June 2018 of Purple
Haze Ltd. During your review, you note the following two independent
situations.
1. Accounts payable is a material account balance comprising 25 large and
150 small accounts payable. Geoff Hunt, the chief executive officer (CEO)
of Purple Haze, presented the junior auditor with a list of accounts
payable making up the balance at 30 June 2018. The junior auditor
decided to randomly select 25 creditors from this list and vouch these
creditors to source documents. The junior auditor indicated in their
working papers that they believed this would provide sufficient
appropriate audit evidence regarding accounts payable.
2. A test of controls regarding 50 sales transactions was undertaken. This
test resulted in three errors. In the planning of the test, a tolerable
deviation rate of 5 per cent was established. The working papers by the
junior auditor noted that none of the errors found was material, either
individually or in aggregate, and therefore the tests of controls
established that the controls were reliable.
REQUIRED
(a) For each of the two situations above, provide advice on the ability of
the audit procedures to provide sufficient appropriate audit
evidence to reduce the audit risk to a tolerable level.
(b) In relation to the accounts payable balance, what details should be
included in the audit documentation regarding the junior auditor’s
procedures? LO 10.6
10.24 MEDIUM You are conducting the audit of Express Player Ltd for the year
ended 30 June 2018. Express Player pays its employees on a weekly
basis. The key control over payroll is that Faith Livingstone, the human
resources manager, reviews and approves the payroll based on the payroll
report produced by the payroll department.
The tolerable rate of deviation for this control has been set at 10 per cent
and you have tested this control using a sample of ten weekly payroll
reports. From this test of controls, eight of the weekly payroll reports were
found to have been appropriately reviewed and approved by Faith. For
the other two payroll reports, the following was found:
REVIEWED AND
APPROVED BY
PAYROLL REPORT FAITH? EXPLANATION
11 February 2018 No Faith did not review

or approve this
payroll report. Faith
stated that this was
because she was
busy handling
employee disputes.
13 May 2018 No Faith was on leave

for a week during
this period and
Express Player’s
CFO reviewed and
approved the payroll
report. Faith reports
to the CFO.
Page 454
REQUIRED
(a) Indicate whether the payroll control can be relied upon and justify
your answer. Explain the effect on the planned substantive tests for
payroll.
(b) Assume that the tolerable rate of deviation for this control has been
set at 0%. Explain the impact this would have (if any) on the planned
substantive tests for payroll. LO 10.6
Sampling for tests of controls: in particular,

attribute sampling
10.25 EASY The audit assistant Pam Priest has been assigned to the audit of
Steel Security Doors Ltd for the year ended 30 June 2018 and has been
asked to complete testing of controls in the revenue and receipts cycle.
Pam reviews the audit program and notes that a sample of 50 sales
invoices is required for testing and that a statistical sampling approach has
been adopted. The required confidence level has been set at 95 per cent
and the tolerable deviation rate is 10 per cent. After requesting access to
copies of invoices, Pam selects 10 sequential invoices from each of the
months of March, April, July, August and December for her testing. The
company processes approximately 1000 invoices per year.
Pam completes her testing and finds only one error. Using a statistical
table (Table 10.4 ), she determines that the actual deviation rate does
not exceed the expected deviation rate. She concludes: ‘This control
appears to be operating as anticipated’.
REQUIRED
Explain whether you agree with Pam’s conclusion, providing reasons for
your answer. LO 10.7
10.26 MEDIUM You have been assigned to the audit of Fantasy Ride Ltd for the
year ended 30 June 2018. You have conducted audit procedures to test
the following controls related to Fantasy Ride’s revenue process.
1. Accounts receivable clerk compares quantity on sales invoice to
shipping documents.
2. Sales manager approves prices on invoices.
You have decided to use a sampling approach based on the following
judgments for each control.
Control Required Tolerable Expected

confidence deviation rate % number of
level % errors
1 95 5 0
2 90 8 2
As a result of your testing, you obtained 4 deviations for control 1, and 2

deviations for control 2.
REQUIRED
(a) Determine the initial sample size to test each of the controls detailed
above.
(b) What should you conclude about each control? Justify your
answer. LO 10.7
Sampling for substantive tests: in particular, dollar-

unit sampling
10.27 MEDIUM Felicity Stone is the auditor for Colossal Ltd for the year ended
30 June 2018. Felicity has decided to use dollar-unit sampling when
auditing accounts receivable.
The value of accounts receivable is $3 762 206. Felicity has established a
tolerable misstatement amount of $23 775 and a sample size of
125 invoices, with a random start of $23 700 and a sample interval of $30
098.
The following is a summary of the first 10 invoices from the total Page 455
population of accounts receivable invoices at Colossal:
INVOICE NUMBER AMOUNT $
A1001 15 000
A1002 7 000
A1003 6 000
A1004 5 000
A1005 12 000
A1006 10 000
A1007 9 000
A1008 17 000
A1009 11 000
A1010 32 000
REQUIRED
(a) Identify which of the first 10 invoices will be selected by Felicity
based on her sampling plan.
(b) Explain why invoice A1010 would always be selected irrespective of
the random starting point. LO 10.8
10.28 MEDIUM You are auditing Furnace Ltd for the year ended 30 June 2018
and have decided to use dollar-unit sampling for accounts receivable. The
sample size will be 25 and a random start of 14 320 has been chosen.
Furnace’s total of the accounts receivable listing has a balance of $400
000, with more than 150 customers. Listed below are the initial 10 items
from the accounts receivable population:
Population Item Recorded Amount ($) Cumulative Value ($)
1 6 857 6 857
2 6 430 13 287
3 820 14 107
4 1 895 16 002
5 3 000 19 002
6 10 500 29 502
7 2 740 32 242
8 4 716 36 958
9 2 769 39 727
10 9 325 49 052
REQUIRED
(a) Explain whether you are more likely to be concerned with testing for
overstatement or understatement for accounts receivable of
Furnace.
(b) Determine which will be the second item to be selected.
(c) You have undertaken your sampling plan of 25 items and your
evaluation of errors reveals that the upper error limit exceeds the
tolerable error. Identify the courses of action you have available.
LO 10.8
Page 456
Other statistical sampling approaches

10.29 EASY You have just attended a staff training course on sampling and the
staff training manager referred to the use of difference estimation.

REQUIRED
Explain what is meant by difference estimation. LO 10.9
Continuous case study
Background information for the continuous case study, Reliable Printers Ltd (RPL), is
contained in the Appendix to this book.
10.30 MEDIUM You have decided to conduct substantive tests of detail on the
accuracy, valuation and allocation assertion for accounts receivable for
RPL, using dollar-unit sampling to decide which invoices to examine. The
value of gross accounts receivable, as indicated in the background
information contained in the Appendix , is $5 313 309 as at 30 June
2018.
The following is a summary of the first 10 invoices from the total population
of accounts receivable invoices outstanding at RPL as at 30 June 2018:
INVOICE NUMBER RECORDED AMOUNT CUMULATIVE

$ VALUE $
19051 20 228 20 228
19052 9 440 29 668
19053 8 091 37 759
19054 6 743 44 502
19055 16 182 60 684
19056 13 845 74 529
19057 12 137 86 666
19058 22 925 109 591
19059 14 934 124 525
19060 10 788 135 313
You have established a tolerable misstatement amount of $103 775 and a

sample size of 66 invoices, with a random start of $31 959 and a sample
interval of $80 587. The total value of the 66 sample items is $1 088 237.
You have examined your total sample of invoices and discovered that the
dollar value of accounts receivable that should have been written off or
provided for in your sample was $40 775.
REQUIRED
(a) Based on the information provided above, identify which of the first
10 invoices will be selected if you design your sample using dollar-
unit sampling.
(b) Given that the $40 775 has not been written off or provided for,
what can you conclude from the sample? LO 10.8

Chap 10 Sampling

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chap 10 Sampling

Uploaded by

Copyright:

Available Formats

Page 421

10.7 Consider sampling approaches to tests of controls, in particular,

ASA 500/ISA 500 Audit Evidence

ASA 530/ISA 530 Audit Sampling

LO 10.1 Deﬁnition and features

how many items to examine

The transformed audit will expand beyond sample-based testing to include

Non-statistical versus statistical audit sampling

Statistical sampling is defined in ASA 530.5(g) (ISA 530.5(g)) as:

an approach to sampling that has the following characteristics:

(a) random selection of the sample items; and

Distinction between statistical and non-statistical sampling

The choice of statistical or non-statistical sampling

Current practice in Australia and the impact of the business

characteristics of a population are inferred from examining specific transactions or account

FIGURE 10.2 Sampling phases

the objectives of the audit test

evaluating the sample results.

When performing substantive procedures, an account balance or class of transactions is

Deﬁning the sampling unit

debtor is a sampling unit).

A sampling technique specific to auditing is a technique called value-weighted selection

1. The auditor’s assessment of the risk of material misstatement This is the

the combination of sampling units actually chosen.

GLOBAL EXAMPLE 10.2 Sampling by systematic selection

Step 2: Give every item in the population an equal chance of selection by

Biases from haphazard sampling

The question (asked) . . . select a number from 1 to 10 and write it in the

Source: Stewart and Dunn, 2000.

Unacceptable sample selection methods

items. ASA 500.A56 (ISA 500.A56) explains the point as follows:

While selective examination of specific items from a class of

LO 10.6 Performing the audit procedures and evaluating

FIGURE 10.3 Projecting misstatements from sample to population

are at an organisational level (rather than a transaction level), such as segregation of

Control risk assessment and tolerable deviation

Qualitative level of control risk Tolerable deviation rate (%)

High Omit test

Determining sample size

where R = reliability factor for desired level of assurance

Number of errors Desired level of assurance

99% 95% 90%

0 4.61 3.00 2.31

1 6.64 4.75 3.89

2 8.41 6.30 5.33

3 10.05 7.76 6.69

4 11.61 9.16 8.00

5 13.11 10.52 9.28

where TDR = tolerable deviation rate.

the tolerable deviation rate

Assume an auditor has speciﬁed a tolerable deviation rate of 7 per cent

Number of expected deviations/misstatements in parenthesis

Expected Tolerable deviation/misstatement rate

0.00 114(0) 76(0) 57(0) 45(0) 38(0) 32(0) 28(0) 25(0

0.25 194(1) 129(1) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1

0.50 194(1) 129(1) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1

0.75 265(2) 129(1) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1

1.00 * 176(2) 96(1) 77(1) 64(1) 55(1) 48(1) 42(1

1.25 * 221(3) 132(2) 77(1) 64(1) 55(1) 48(1) 42(1

1.50 * * 132(2) 105(2) 64(1) 55(1) 48(1) 42(1

1.75 * * 166(3) 105(2) 88(2) 55(1) 48(1) 42(1

2.00 * * 198(4) 132(3) 88(2) 75(2) 48(1) 42(1

2.25 * * * 132(3) 88(2) 75(2) 65(2) 42(1

2.50 * * * 158(4) 110(3) 75(2) 65(2) 58(2