11/16/2023

AC 306: LECTURE 2

AUDIT SAMPLING

EC 1

To be discussed

• Definition and features

• Sampling criteria
• Statistical vs. non-sampling sampling
• Probabilistic vs non-probabilistic
• Basic requirements of all audit samples
• Factors leading to non-representative sample
• Sampling for tests of controls, attribute sampling
• Nonstatistical and statistical audit sampling for substantive tests of
details
• Procedures that may not involve audit sampling

EC 2

1
 11/16/2023

Definition and features

Objectives of sampling
• To gather sufficient appropriate evidence
• To obtain a sample whose characteristics are the same as those of
the population. (aka ‘representative’ sample).
Why sample?
• Impracticability of auditing 100%
• To minimise cost of the audit

EC 3

Definition and features

✓Audit sampling: the application of an audit procedure to less than

100% of the items within a population to obtain audit evidence
about particular characteristics of the population.
✓Audit sampling is important because it provides information on:
i. How many items to examine
ii. Which items to select
iii. How sample results are evaluated and extrapolated to the population in
order to tell something about the population (e.g. level of misstatement).
✓Population – The entire set of data from which a sample is
selected and about which the auditor wishes to draw conclusions.

EC 4

2
 11/16/2023

Definition and features

✓ Stratification – The process of dividing a population into

subpopulations, each of which is a group of sampling units which
have similar characteristics (often monetary value).
✓ Sampling risk – The risk that the auditor’s conclusion based on a
sample may be different from the conclusion if the entire
population were subjected to the same audit procedure. It can lead
to 2 types of errors:
i. In the case of a ToC, that controls are more effective than they actually
are, or in the case of a ToD, that a MM does not exist when in fact it
does.
ii. In the case of a ToC, that controls are less effective than they actually are,
or in the case of a ToD, that a MM exists when in fact it does not.
✓ Anomaly – A misstatement or deviation that is demonstrably not
representative of misstatements or deviations in a population.
✓ Sampling unit – The individualEC items constituting a population5.

Sampling criteria

• When sampling, the auditor identifies a particular

characteristic of the population to focus upon.
i. For tests of control, the characteristic of interest is the rate
of deviation from an internal control policy or procedure.
ii. For substantive tests, the characteristic of interest is
monetary misstatement in the balance.
• Various means of gathering audit evidence
i. 100% examination: this is not a sampling method.
ii. Selecting specific items: e.g. high value or high risk — this
is not a sampling method. Items selected will not
necessarily be representative of the population.

EC 6

3
 11/16/2023

Statistical vs. non-sampling sampling

Statistical sampling: an approach to sampling that has the

following characteristics:
i. Random sample selection
ii. Use of probability theory to evaluate sample results

- Use of mathematical techniques to calculate formal

statistical results
- Ideal for large population items with ready access to each
item.
• Major advantage is defensibility, thorough quantification of
sampling risk.

EC 7

Statistical vs. non-sampling sampling

Non-statistical sampling: sampling approaches that do not have all
the characteristics of statistical sampling.
• Is a more subjective approach to inference: mathematical
techniques not consistently used to determine sample size, selecting
sample items, or evaluating results
• Relies heavily on the auditor’s professional judgment
• Ideal for smaller populations, or populations with items not so
readily accessible by the auditor for sampling
• Does not quantify sampling risk

EC 8

4
 11/16/2023

Statistical and non-sampling sampling

Parts of statistical/non-statistical sampling methods
i. Sample size determination (deciding how many items to
select).
ii. Sample selection (deciding how to select the items). Methods
are:
❑ Probabilistic: each population item has a known chance of
being selected.
❑ It is the required method for statistical evaluation; non-
statistical evaluation is acceptable.
❑ Non-probabilistic (judgmental): auditor decides which items to
select.
iii. Performing the tests.
iv. Sample evaluation: drawing conclusions based on audit tests.
EC 9

Probabilistic vs non-probabilistic
• Probabilistic sampling: random selection
Random sample: every possible combination of items in
population have equal chance of being selected.
Common random selection methods:
i. Use of random number tables:
Will need to consider discards, documentation and replacement v. non-replacement
sampling.
ii. Use of computer software.
iii. Systematic sampling:
Advantage: it is easy to use, sample can be drawn quickly, and
documentation is easy.
Disadvantage: may cause bias (not preferred by some audit firms).

EC 11

5
 11/16/2023

Probabilistic vs non-probabilistic

• Non-probabilistic sampling: non-random selection

Common approaches are:
i. Judgmental methods: -
NB. random sampling may not give a representative sample when
sample sizes are small.
ii. Haphazard selection: -
Has no regard to size, source, or other distinguishing characteristics.
Pseudo random; but difficult to remain unbiased.
iii. Block sampling (Cluster sampling):
The selection of several items sequentially.
Selection of blocks, then all items in block are chosen.
Is acceptable only if a reasonable number of blocks are used.

EC 12

Basic requirements of all audit samples

Whenever an auditor uses audit sampling (statistical or non-statistical)
the following requirements apply:
i. Planning and design: 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.
ii. Selection: Sample items are selected in such a way that the sample
can be expected to be representative of the population.
iii. Performing the procedure and evaluating 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.
EC 12

6
 11/16/2023

Basic requirements of all audit samples

Q.How can the auditor improve representativeness?
• A. The likelihood of a sample being a representative one can be
increased by using care in its design, selection, and evaluation.
Since the need is to gather sufficient appropriate evidence,
consider:
• Is the sample large enough to represent the population?
• Is sampling relevant in the circumstances of this population?
• Are the selection procedures representative enough to make the
sample suitable for assessing the reliability of the population?

EC 13

Planning and designing the sample

Auditor must consider:

• Objectives of the audit test (usually related to an audit
assertion of interest)
• Population from which to sample
• Possible use of stratification
• Definition of the sampling unit.

EC 14

7
 11/16/2023

Planning and designing the sample

Auditor must consider:
1. Objectives of the audit test (usually related to an audit assertion
of interest)
✓ Once the audit objective is specified, such as reliance on
controls or misstatement of account balance, the auditor
must consider what conditions would constitute an error.
✓ The auditor must ensure that the population from which the
sample is to be selected is complete and appropriate to the
audit objective.
2. Population from which to sample.

EC 15

Planning and designing the sample

3. Possible use of stratification
✓ Stratification: occurs when the auditor divides the population
into a series of sub-populations, each of which has an
identifying characteristic, such as dollar value
✓ Can assist with audit efficiency as it allows the auditor to
reduce the sample size by reducing variability without
increasing the sampling risk.
✓ Can direct auditor’s attention to areas of audit interest,
especially risky or material items.
4. Definition of the sampling unit - Transactions or balances making
up the account balance

EC 16

8
 11/16/2023

Determining sample size

• Sample size is affected by the degree of sampling risk the auditor is
willing to accept.
• Auditor's major consideration in determining sample size is
whether, given expected results from examining sample, sampling
risk will be reduced to an acceptably low level

EC 17

Determining sample size

Factors that influence sample size for tests of controls- Appendix 2
to ISA 530 outlines the factors that influence sample size for tests of
controls as follows:
i. The extent to which the auditor’s risk assessment takes into
account relevant controls (control risk assessment).
ii. The tolerable rate of deviation – highest amount of error that an
auditor can accept in a sample.
iii. The expected rate of deviation- rate of mistakes anticipated by
an auditor based on his/her prior knowledge or assumptions.
iv. The auditor’s desired level of assurance
v. The number of sampling units in the population.

EC 18

9
 11/16/2023

Determining sample size

Factors that influence sample size for substantive testing. Appendix 3 to
ISA 530 outlines the factors that influence sample size for substantive
testing as follows:
1. The auditor’s assessment of risk of material misstatement
2. The use of other substantive procedures directed at the same assertion
3. The auditor’s desired level of assurance that actual misstatement is not
greater than tolerable misstatement
4. The tolerable misstatement - If errors surpass the limit, the auditor may
need to investigate and request corrections.
5. The amount of misstatement the auditor expects to find in the
population (expected misstatement)
6. Stratification
7. The number of sampling units in the population.
EC 19

Selecting the sample

• To draw conclusions about population or stratum, the sample
needs to be typical of characteristics of population or stratum.
• Sample needs to be selected without bias so that all sampling
units in the population or stratum have a chance of selection.
• Common sampling techniques are:
✓ Random selection — random number generation
✓ Systematic selection
✓ Haphazard selection — select without conscious bias.

EC 20

10
 11/16/2023

Selecting the sample

Steps in systematic selection
• For example, suppose the sample size is 20 and the number of
items in the population is :
▪ Step 1: Calculate the sample interval, (say 500 i.e. 10,000/20)
▪ Step 2: Give every item in population chance of selection by
choosing a random number (random start) within range of 1
and sampling interval (in this example, 500), e.g. 217.
▪ Step 3: Continue to add sampling interval to random start, and
identify items to be sampled, e.g. item nos. 217, 717, , 9717.

EC 21

Selecting the sample

Unacceptable sample selection methods
• Block selection: the auditor selects all items of a specified type
processed on a particular day, week or month.
• Judgmental selection (based on sample item characteristics): the
auditor selects large or unusual items from the population or uses
some other judgmental criterion for selection. This method has a
conscious bias and cannot be considered representative.

EC 22

11
 11/16/2023

Performing the audit procedures

Evaluating sample results
• To evaluate sample results, auditor determines the level of
misstatement found in sample and directly projects this
misstatement to relevant population.
• For example: sample 20%, find misstatement of $ Therefore projected
misstatement = $ ($10 000/20%).
• Projected misstatement is then compared with tolerable
misstatement for the audit procedure to determine if characteristic
of interest can be accepted or rejected.
• Auditor should consider both the nature and cause of any
misstatement or deviations identified.

EC 23

Factors leading to non-representative sample:

• Non-sampling error- occurs when the audit does not uncover
exceptions in the sample. A.k.a. non-sampling risk, can be
eliminated by careful design of audit procedures and proper
supervision and instruction of audit staff. Caused by:
- auditor’s failure to recognise exceptions (e.g. exhaustion, boredom, lack of
understanding, not knowing what to look for, etc.)
- inappropriate or ineffective audit procedures (e.g. wrong test method used)
Sampling error- An inherent part of sampling resulting from testing
less than 100%. Sampling error is always there even with zero non-
sampling error (there is always a chance that a sample is not
representative). A.k.a. sampling risk, can be reduced by
- increasing sample size, or
- using appropriate method of selecting sample items from the population
(e.g. Probabilistic techniques).

EC 24

12
 11/16/2023

Sampling for tests of controls, attribute sampling

• Audit sampling is useful for tests of controls, especially involving
inspection of source documentation for specific attributes such as
evidence of authorisation (attribute sampling).
• Involves examination of documents for particular attributes related
to controls (e.g. authorisation).
• Results of attribute sampling can be used to support or refute an
initial assessment of control risk.

EC 25

Sampling for tests of controls, attribute sampling

In planning and designing sample for tests of controls, auditor should
consider:
• Audit objectives (assertions of audit interest)
• Control risk assessment and tolerable deviation rate
• Allowable risk of over-reliance — allowable risk of assessing
control risk too low
• Expected error — amount of error the auditor expects to find in
the population.

EC 26

13
 11/16/2023

Sampling for tests of controls, attribute sampling

Audit objectives
• A sampling technique that is suitable for tests of controls is
attribute sampling. The following steps are necessary in
considering the relationship between the sample and the objective
of the test:
i. Identify relevant control objectives, policies and procedures
that are relevant to restricting substantive tests of the related
account balances.
ii. Identify population and sample unit.
iii. Define the characteristic of interest – so that an attribute
either exists or does not exist which means that a control
activity has either been complied with or not complied with.
EC 27

Sampling for tests of controls, attribute sampling

Selection of sample items for tests of controls
• After 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 generally apply to both tests of controls and
substantive tests.
• However, stratification is not usually applicable to tests of controls.
• Systematic selection 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.

EC 28

14
 11/16/2023

Nonstatistical and Statistical Audit Sampling for

Substantive Tests of Details
Considers the following factors
i. Assessed risks of material misstatement
ii. Characteristics of the population
iii. Tolerable misstatement
iv. Expected misstatement
v. Audit risk and sampling risk (that is, that actual misstatement exceeds tolerable
misstatement)
vi. Audit evidence obtained from other substantive procedures related to the
same assertion
vii. Selection of a sample that can be expected to be representative
viii. Projection of the sample results to the population
ix. Consideration of an allowance for sampling risk (precision

EC 29

Nonstatistical and Statistical Audit Sampling for

Substantive Tests of Details
Considering the Completeness of the Population
Identifying Individually Significant Items - the auditor uses judgment
to identify the items and test them separately from the remainder
Defining the Sampling Unit - t might be a customer account balance,
an individual transaction, or an individual entry within a transaction. t
depends on the audit objective and the nature of the audit procedures
to be applied.
Choosing an Audit Sampling Technique – statistical or non statistical

EC 30

15
 11/16/2023

Factors Influencing Sample Sizes for a Substantive Test

of Details in Sample Planning
Related Factor for
Small large
Factor Substantive
Sample Size sample size
Sample Planning
Allowable risk of incorrect
Assessment of inherent risk Low High
acceptance
Allowable risk of incorrect
Assessment of control risk Low High
acceptance
Assessment of risk related to
other substantive Allowable risk of incorrect
Low High
procedures directed at the same acceptance
assertion
Measure of tolerable
misstatement for a specific Larger Smaller Tolerable misstatement
account
Expected size and frequency of
misstatements, Assessment of population
Smaller Larger
or the estimated variance of the characteristics
population
Number of items in the
Virtually no effect on sample size unless population
population EC 31

Procedures That May Not Involve Audit Sampling

1. Inquiry and Observation

• Interviewing management and employees
• Obtaining an understanding of the internal controls
• Observing the behavior of personnel and the functioning of business
operations
• Observing cash-handling activities
• Observing the operation of controls
• Performing walkthrough procedures
• Observing the existence of land and buildings
• Obtaining written representations from management
2. Analytical Procedures - do not result in projecting
3. Procedures Applied to Every Item in a Population
EC 32

16
 11/16/2023

Procedures That May Not Involve Audit Sampling

4. Some Tests of Controls May Not Involve Audit Sampling - For

example, tests of automated application controls or determning
appropriate segregation of duties
5. Tests of Controls When Extrapolation Is Not Intended - auditor does
not intend to extend the resulting conclusion to the remaining items.
E.g. trace several transactions through an entity's accounting system to
obtain an understanding of the design of the entity's internal control
6. Untested balances due to for example low RoMM
7. Test of automated IT controls - when IT general controls are tested
and determined to be effective, a single test of an automated control
for each type of control operation may be sufficient

EC 33

Assignment & food for thought

• Can a statistical sampling approach use both
probabilistic and non-probabilistic techniques? With no
fewer than five credible references, argue why or why
not.

EC 11

17
 11/16/2023

THE END!
THANK YOU!

EC 35

18