population can be divided into sampling units in substantive procedures provide audit evidence

a variety of ways. For example, if the auditor’s regarding the same financial statement
objective, were to test the validity of debtors, the assertion. The more other substantive
sampling unit could be defined as customer procedures lower the detection risk for a
balances or individual customer invoices. particular assertion and therefore the lower the
Auditors defined the sampling unit in order to reliance on the results of the substantive
obtain an efficient and effective sample to procedure using audit sampling, the higher the
achieve to particular audit objectives. acceptable sampling risk relating to the
sampling procedure and consequently, the
The population is physically represented by
smaller the sample size.
some form of record. For instance, sales invoices,
entries in a sales journal or summary entries in TOLERABLE ERROR
a ledger may represent total sales. The auditor
Tolerable error is the maximum error in the
selects sampling units from this physical
population that auditors would be willing to
representation and so must confirm that all
accept and still conclude that the result from the
sample units from this record are included in the
sample has achieved the audit objective.
entire population. If the physical representation
Tolerable error is considered during the planning
differs from the actual population, the auditor
stage and, for substantive procedures, is related
might make erroneous conclusions about the
to the auditors’ judgement about materiality. The
population. A simple example of this is testing a
smaller the tolerable error, the greater the
depreciation schedule where a page of the
sample size needs to be.
schedule is missing.
In tests of control, the tolerable error is the
Therefore, the auditor should be careful to
maximum rate of deviation from a prescribed
determine that the records used to draw the
control procedure that auditors would be willing
sample are complete and reflect the actual
to accept and still conclude that the preliminary
population being tested. One means of doing this
assessment of control risk is valid. In
is comparing different records and reconciling
substantive procedures, the tolerable error is the
differences found.
maximum monetary error in account balance or
RISK AND ASSURANCE a class of transaction that auditors would be
willing to accept so that when the results of all
When determining sample sizes, auditor should
audit procedures are considered, auditors are
consider the risk involve in the audit
able to conclude, with reasonable assurance,
Sampling risk arises from the possibility that that the financial statements are not materially
the auditors’ conclusion, based on sample, may misstated.
be different from the conclusion that would be
EXPECTED ERROR IN THE POPULATION
reached if the entire population were subjected
to the same audit procedure. Sampling risk for If auditor expect errors to be present in the
substantive procedures is one form of detection population, a larger sample than when no error is
risk, which is the risk that the auditors’ the expected ordinarily needs to be examined to
procedures do not detect a misstatement that conclude that the actual error in the population
exists in an account balance or class of is not greater than the planned tolerable error.
transactions. Smaller sample sizes are justified when the
population is expected to be error free.
Sample size is affected by the level of sampling
risk auditors are willing to accept from the In determining the expected error in a
results of the sample. The level of acceptable population, auditors would consider such
sampling risk depends upon the importance of matters as error levels identified in previous
the results of the audit procedure involving audits, changes in the entity’s procedures and
sampling to the auditors’ conclusions. The evidence available from other procedures,
greater the reliance on the results, the lower the including tests of control.
sampling risk auditors are willing to accept and
STRATIFICATION
the greater the sample size need to be.
To assist in the efficient and effective design of
The extent of reliance on the results of the
the sample, stratification may be appropriate.
procedure is related to the extent to which other
Stratification is the process of dividing a 3. JUDGMENTAL SELETION
population into sub-population, each of which is
Judgmental selection, which may be an
a group of sampling units, which have similar
acceptable alternative to random selection,
characteristics (often monetary value). The
provides the auditors an attempt to draw a
strata need to be explicitly defined so that each
representative sample from the entire
sampling unit can belong to only on stratum.
population with no intention to either include or
This process reduces the variability of the items
exclude specific units. When auditors use this
within each stratum. Stratification enables
method, care needs to be taken to guard against
auditors to direct audit efforts towards the items
making a selection that is biased, for example,
which, for example, contain the greatest
towards items which are easily located, as they
potential monetary error. For example, auditors
may not be representative.
may direct attention to larger value items for
debtors to detect overstated material 4. STRATIFIED RANDOM SAMPLING
misstatements. Consequently, stratification
may result in smaller sample size. As explained previously, stratification is the
process of dividing a population into sub-
SAMPLE SELECTION METHOD populations, each of which is a group of
sampling units, which have similar
According to PSA 530, auditors should select
characteristics (often monetary value).
sample items in such a way that the sample can
be expected to be representative of the SAMPLE SIZE
population in respect of the characteristics
being tested. For a sample to be representative of When determining the sample sizes, auditors
the population, all items in the population are should consider sampling risk, tolerable error,
required to have an equal or known probability of and the expected error. Table 9.2 and9.3 shows
being selected. While there are number of examples of some factors of some factors
selection methods, four methods are commonly affecting the sample size.
used are set out below. EVALUATION OF SAMPLE RESULTS
1. RANDOM NUMBER SELECTION Having carried out, on each sample item, those
Under this method, the auditor ensures that all audit procedures that are appropriate to the
items in the population have an equal chance of particular audit objective, auditors should:
selection, for example, by use of random number  Analyze any errors detected in the sample
tables.
 Draw inferences for the population as a
2. SYSTEMATIC SELECTION whole

Systematic selection involves selecting items ANALYSIS OF ERRORS IN THE SAMPLE

using a constant interval between selections, the
Before analyzing the errors detected in the
first interval having a random start. The interval
sample, auditors would first determine that an
might be based on a certain number of items (for
item in question is in fact an error. In designing
example, every 20th voucher number) or on
the sample, auditors define those conditions
monetary totals (for example, every P1,000
that constitute an error by reference to the audit
increase in the cumulative value of the
objectives. For example, in a substantive
population). When using systematic selection,
procedure relating to the recording of debtors, a
auditors would determine that the population is
misposting between customer accounts does
not structured in such a manner that the
not affect the total debtors. Therefore, it may be
sampling interval corresponds with a particular
inappropriate to consider this is an error in
pattern in the population. For example, if in a
evaluating the sample results of this particular
population of branch sales, a particular branch’s
procedure, even though it may have an effect on
sales occur only as every 100th item and the
other areas of the audit such as the assessment
sampling interval selected is 50, the result would
of doubtful accounts.
be that, or none, of the sales of that particular
branch would have been selected. When the expected audit evidence regarding a
specific sample item cannot be obtained,
auditors may be able to obtain sufficient
appropriate audit evidence through performing
alternative procedures. For example, if a positive substantive procedures is net of adjustments
debtor confirmation has been requested and no made by the entity. When the projected error
reply was received, auditors may be able to exceeds tolerable error, auditors re-assess the
obtain sufficient appropriate audit evidence that sampling risk and if that risk is unacceptable,
the debtors is valid by reviewing subsequent consider extending the audit procedure or
payments from the customer. If auditors do not, perform alternative audit procedures, either of
or are unable to, perform satisfactory alternative which may result in proposing an adjustment to
procedures or if the procedures performed do not the financial statements.
enable auditors to obtain sufficient appropriate
Table 9.2 Examples of Factors Influencing
audit evidence then the item would be treated as
Sample Sizes for Tests of Control
an error.
Conditions Leading to
Auditors would also consider the qualitative
Factor Smaller Larger Sample
aspects of the errors. These include the nature
Sample Size Size
and cause of the error and the possible effect of
Preliminary Higher (1) Lower
the error on other phases as an error. assessment of
In analyzing the errors discovered, auditors may control risk
observe that many have a common feature. For Tolerable Error Higher Lower
example, type of transaction, location, Acceptable risk Higher Lower
of over reliance
production line or period of time. In such
Expected error Lower Higher (2)
circumstances, auditors may decide to identify
No. of items in Virtually no effect on sample
all items in the population which possess the
population size unless population is small
common feature, thereby producing a sub-
population, and extent audit procedures in this
area. Auditors would then perform a separate (1) A high preliminary assessment of control
analysis based on the items examined for each risk may result in a decision NOT TO
sub-population. PERFORM TESTS OF CONTROL
(2) High expected deviation rates may result
INFERENCES TO BE DRAWN FOR THE POPULATION
in a decision NOT TO PERFORM TESTS OF
AS A WHOLE
CONTROL
This covers the projection of errors and the re-
Preliminary assessment of control risk is
assessment of sampling risk. Auditors project
INVERSELY related to sample size. Therefore, a
the error results of the sample to the population
DECREASE in preliminary assessment of control
from which the sample was selected in order to risk will case the sample size to INCREASE.
form a conclusion about the possible level of
error in the population as a whole. The projection Tolerable error is INVERSELY related to sample
of the error results of the sample to the size. The smaller the tolerable error, the larger the
population as a whole involves estimating the sample size.
probable error results, auditors would ensure
Acceptable risk of over reliance is INVERSELY
that the method of projection is consistent with
related to sample size.
the method used to select the sampling unit.
This is in addition to considering the qualitative Expected error has a DIRECT EFFECT on the
aspects of the errors found. When the population sample size. The larger the expected deviation
has been divided into sub-populations, the rate, the larger would be the sample size.
projection of errors is done separately for each
sub-population and the results are combined. Table 9.2 Examples of Factors Influencing
Sample Sizes for Tests of Control
Auditors would consider whether errors in the
population might exceed the tolerable error. To Conditions Leading to
accomplish this, auditors compare the projected Factor Smaller Larger
population error to the tolerable error taking into Sample Size Sample Size
Assessment Lower control Higher
account the results of other audit procedures
of control risk control risk
relevant to the specific control or financial
risk,
statement assertion. The projected population supported by
error used for this comparison in the case of