Paper F8: Audit and assurance (International)
146 Go to www.emilewoolfpublishing.com for Q/As, Notes & Study Guides © EWP
For example, if the objective of the audit testing is to confirm the accuracy of trade
receivable balances, the population should include all trade receivables balances as
at a specified date. Alternatively, the audit objective might be more limited in scope
– perhaps to confirm the accuracy of those trade receivables balances in excess of
$500,000. The population should then consist of all receivables balances over
$500,000.
Each individual item in the population (trade receivables in the example above) is
referred to as a sampling unit. It is important that all sampling units should be
homogeneous, (have the same characteristics). In the example of the trade
receivables balances above, the auditor would have to ensure that all the individual
balances have been processed by the same accounting methods.
If this is not the case, the population may not be sufficiently homogeneous to allow
a valid sample to be taken. However, it may be possible to turn a non-homogeneous
population into two or more homogeneous populations and then sample from each
of these.
For example, suppose that trade receivables balances have been processed through
two different accounting systems (perhaps a computerised accounting system and a
manual, paper-based system). The entire population of receivables could be divided
into two segments (or strata) and a sample could be selected for testing from each
segment. This technique is known as stratified sampling.
The sample selection method
ISA 530 requires auditors to select a sample in such a way that each item in the
population (each ‘sampling unit’) has an equal chance of being selected.
A wide range of sample selection methods is available to the auditor:
Random sampling: All items in the population have an equal chance of
selection. This is typically achieved by the use of random numbers to select
items for testing.
Systematic sampling: With systematic sampling, a random starting point is
chosen from the population and then items are selected with a standard gap
between them (for example, every 10th item). For example, suppose that a
sample will be 10% of the items in a population and the items in the population
can be arranged in a sequence, such as listed in invoice number order, or account
number order or date order. A systematic sample would be to select one of the
first 10 items in the list at random, and then to select every 10th item in the list
for testing in order to obtain the 10% sample.
Haphazard sampling: The auditor selects the sample on an arbitrary basis, for
example, choosing any 100 invoices from a file. This is not a scientifically valid
method and the resulting sample may contain a degree of bias. It is therefore not
recommended for use with statistical sampling techniques.