AUDIT SAMPLING Sampling Risk and Its Possible Effects

Involves the application of audit procedures to Sampling Risk refers to the risk that the
less than 100% of items within a population of auditor’s conclusion based on a sample may be
audit relevance such that all sampling units have different from the conclusion if the entire
a chance of selection. population were subjected to the same audit
procedure.
Audit sampling is designed to enable
conclusions to be drawn about an entire The mathematical compliment of sampling risk
population on the basis of testing a sample is known as confidence level or reliability level.
drawn from it.
As a TEST OF CONTROL procedures
NOTE: Audit sampling can be eliminated.
Risk of Risk of Affects Type 1
Objection of the Auditor when Sampling is assessing UNDER auditor’s risk
Used Control reliance efficiency and
Risk TOO Alpha
 Design and select audit sample HIGH risk
 Perform audit procedures on the Risk of Risk of Affects Type 2
sample items assessing OVER auditor’s risk
 Evaluate the results from the sample control effectiveness and
 Provide an appropriate bases to draw risk TOO Beta
conclusions about the population from LOW risk
which the sample is drawn

Audit Sampling As a SUBSTANTIVE TEST procedures

Requirements Conclusion Risk of Affects Type 1 risk and

Reached incorrect auditor’s Alpha risk
Test of The control to Whether rejection efficiency
Controls be tested controls are Risk of Affects Type 2 risk and
leaves effective or incorrect auditor’s Beta risk
documentary ineffective. acceptance effectiveness
evidence or
audit trail.
NOTE: The Auditor is more concerned with
Substantive The auditor Whether
Beta risk because the provision of reasonable
Testing applies test of the account
details (either balance assurance depends on it.
test of tested is Managing Sampling Risks
transactions fairly stated
and events or or Examination of the entire population (100%) or
test of additional selecting specific items would ELIMINATE
balance is procedures sampling risks.
used are required
to be Only exposed to sampling risk when applying
performed. Audit Sampling

Auditors control sampling risk by:

RISKS INVOLVED WHEN GATHERING EVIDENCE
Increasing the sample size
Using an appropriate sample selection method similar characteristics to each other,
but different characteristics from
Methods of Selecting Samples items elsewhere in the population.
 Random-number selection
- Under this method, every sampling Non-Sampling Risks
unit has the same probability of - Pertains to the risk that the auditor
being selected reaches an erroneous conclusion for
- Applied through random number any reason not related to sampling
generators, for example, random risk.
number tables or computer - It is also defined as the possibility
programs that the auditor does not recognize
- Documents must be pre-numbered misstatements or deviations
 Systematics Selection included in the sample for what
- Applied through calculation of they are.
sampling interval dividing the - It arises from factors that cause the
population with the sample size. auditor to reach an erroneous
- Using the calculated sampling conclusion for any reason not
interval, evert nth item is selected related to the size of the sample.
after a random start (a random Examples:
starting point gives all items in the
 Inappropriate selection of audit
population an equal opportunity to
procedures
be selected.
 Misinterpretation of results of audit
- Must be applied on a surprise basis.
tests
 Haphazard
 Failure to recognize misstatements or
- Auditor selects the sample without
deviations in sample tested
following a structured technique.
Managing Non-Sampling Risks
- Although no structured technique is
used, the auditor would  Adequate Planning
nonetheless avoid any conscious  Proper direction, supervision and
bias or predictability (for example, review of audit team
avoiding difficult to locate items, or  Adherence to firm’s quality control
always choosing or avoiding the first standards
or last entries on a page) and thus
attempt to ensure that all items in GENERAL APPROACHES TO AUDIT SAMPLING
the population have a chance of Statistical sampling pertains to an approach to
selection. sampling that has the following characteristics:
 Block or Cluster Selection
- It involves selecting a block/s of  Random selection of the sample items;
contiguous items from within the and
population.  The use of probability theory to
- It cannot ordinarily be used in audit evaluate sample results, including
sampling because most populations measurement of sampling risk.
are structured such that items in
sequence can be expected to hav e Statistical sampling helps the auditor to:
  Design an efficient sample that would - Population is the class of
meet given objectives transactions being tested. This
would involve defining the period
 Measure the sufficiency of the evidence
covered by the test, defining the
 Evaluate the sample result sampling unit, and consideration of
the completeness of the population.
Statistical sampling applies the laws of 3. Define the attribute and deviation
probability in determining the sample size. This conditions
allows the auditor to quantify sampling risk to - An auditor should identify
limit it to an acceptable level. characteristics or attributes that
Non-statistical sampling is any approach to would indicate operations of the
sampling that does not have the characteristics internal control is operating
of statistical sampling. effectively. A deviation is the
absence of an attribute.
Non-statistical sampling exclusively relies on 4. Determine the sample size
judgment to determine the sample size and - Sample size is determined by
evaluate sample results. considering the following factors:
SAMPLING PLANS  Acceptable level of sampling risk
 Risk of assessing control risk too low
- Attribute Sampling Plan  Tolerable deviation rate
- Variable Sampling Plan  Expected deviation rate
Attribute Sampling Plan (Used in Test of Summary of relationship of the different
Controls) factors to be considered in determining sample
size
 This is a sampling plan used to estimate
the rate or frequency of
occurrence of a certain characteristic in
a population

 This is performed during test of controls

to estimate the rate of deviations
from prescribed internal control policies
or procedures.
NOTE:
 Generally used when there is a trail of
In risk of assessing too low is LOW therefore the
documentary evidence.
sample size will gather is LARGE. This is because
The expected deviation rate and sample size a lower error rate requires more data to detect
have a direct relationship because: a true effect or relationship with a higher
degree of confidence, vice versa.
Steps in Attribute Sampling Plan
Tolerable deviation rate refers to the maximum
1. Determine the objective of the test rate of errors or deviations you’re willing to
- The auditor should clearly state accept in the population. Acceptable sampling
what is to be accomplished or the risk refers to the risk of making a wrong
specific objectives to be attained.
2. Define the population
conclusion about the population based on the In case of properly voided, unused, or
sample. inapplicable documents selected as samples
can be replaced.
The tolerable deviation rate and acceptable
sampling risk have an inverse relationship in In case of missing documents, it should be
selecting sample size because: considered as a deviation since the auditor was
unable to determine whether the control has
 A higher tolerable deviation rate means
been properly performed.
you're willing to accept more errors in
the population, so you need a smaller If in the early stage of the sampling process
sample size to make your conclusions. enough deviations had been found, the auditor
need not continue the tests.
 A lower tolerable deviation rate means
you want to detect even small errors, so 7. Evaluate the sample results
you need a larger sample size to reduce - Once audit procedures have been
the risk of missing those errors (lower performed on all samples, the
sampling risk). sample result must be evaluated.
- Procedures performed include:
 Determining the sample
The expected deviation rate and sample size deviation rate
have a direct relationship because:  Determining the maximum
population deviation rate
 If the population is more varied (higher (achieved upper deviation limit)
expected deviation), you need more data and the allowance for sampling
(larger sample) to make sure your results risk (achieved precision)
are accurate. Conversely, if the population  Consider the qualitative aspects
is less varied (lower expected deviation), a of the deviation.
smaller sample size can still give reliable  Reach an overall conclusion.
results.
8. Document the sampling procedures
5. Determine the method of selecting the - Each of the prior steps, as well as
sample the basis for overall conclusions,
- The sample to be selected should should be documented in the
be representative of the population. working papers.
All items in the population should
have an opportunity to be selected. OTHER ATTRIBUTES SAMPLING TECHNIQUES
6. Test the sample item
Sequential sampling
- The auditor shall apply the
appropriate audit procedures to all  The sampling plan is performed in
items in the sample to determine if several steps
there are any deviations from the
 Every after each step, the auditor
prescribed controls procedures.
decides whether to stop or to go on to
- Deviations may be an isolated or
the next step.
recurring type of occurrence
  This method is used when the auditor  If statistical is used, probability-
expects zero or very few deviations proportional-to-size sampling (PPS) or
within the population. classical variable techniques are
appropriate.
Discovery sampling
4. Determine the sample size
 This method is used when the auditor
- After determining the items to be
expects near zero deviations within the
tested, sample size should be
population.
determined.
 If the auditor discovers at least one - Sample size is determined by
deviation, no reliance would be made to considering the following factors:
the internal controls.
• Acceptable level of risk
VARIABLE SAMPLING (Used in Substantive
• Variation within the population
Testing)
• Tolerable misstatement or error
A. This is a sampling plan used to estimate
numerical measurement of a • Expected misstatement or error
population such as peso value.
Summary of relationship of the different
B. This is generally performed in factors to be considered in determining sample
substantive testing to estimate the size
amount of misstatements in the
financial statements of the entity under
audit.

Steps in Variable Sampling Plan

1. Determine the objective of the

substantive procedure
- The auditor’s objective is to test the
reasonableness of a recorded NOTE:
account balance in the financial The risk of incorrect acceptance has an inverse
statements. relationship with sample size because:
2. Define the population and sampling
unit Lower Risk Requires More Evidence (Larger
- Population is the class of Sample): To reduce the risk of incorrect
transactions being tested. This acceptance, you need to gather more
would involve defining the period information to make a more accurate decision.
covered by the test, defining the A larger sample size allows you to better detect
sampling unit, and consideration of defects or issues within the population,
the completeness of the population. increasing the chance of identifying a bad lot.
3. Selecting an audit sampling technique With more data points, your conclusion is more
reliable and less likely to be wrong. Higher Risk
 The auditor may use either statistical or Allows for Fewer Samples: If you're willing to
non-statistical sampling. accept a higher risk of incorrect acceptance (i.e.,
you're more tolerant of errors), you don't need
as much data. A smaller sample size might be  If the audit procedure is not
sufficient, and you reduce the cost or effort applicable to the selected item,
involved in sampling. the auditor shall perform the
procedure on a replacement item
(e.g., the item is properly voided)
5. Determine the method of selecting the  If the auditor is unable to apply
sample the designed audit procedures, or
- When selecting a sample for suitable alternative procedures,
substantive testing, the auditor may to a selected item the auditor
use any one of the sample selection shall treat that item as a
methods discussed in attribute misstatement, in the case of tests
sampling plan. of details (e.g., missing
documents)
Stratification 7. Evaluate the sample results
• This method involves dividing a - Once audit procedures have been
population into subpopulations, each of performed on all samples, the sample
which is a group of sampling units which - result must be evaluated. This
have similar characteristics (often step would include the following:
monetary value).  Project the misstatements to
the population
• The strata must be explicitly defined so  Considering sampling risk
that each sampling unit can belong to  Considering qualitative
only one stratum. information
• This method enables the auditor to  Reaching an overall conclusion
direct his efforts towards the items he 8. Document the sampling procedures
considers would potentially contain the
greater monetary error.
OTHER VARIABLE SAMPLING PLAN
• It decreases the effect of variance in the TECHNIQUES
population.
Classical Variable Sampling Techniques
Value weighted selection
Mean-per-unit Estimation is a classical variable
• Under this method, the probability of sampling technique that projects the sample
an item to be selected is directly average to the total population by multiplying
proportional to the monetary value of the sample average by the number of items in
such item. population.
• Also, each peso is treated as one Difference Estimation is a classical variable
sampling unit. sampling technique that uses the average
6. Test the sample item difference between audited amounts and
- The auditor shall perform audit individual recorded amounts to estimate the
procedures appropriate to the total audited amount of a population and an
particular audit objective on each allowance for sampling risk.
item selected.
Difference estimation is more appropriate when
there is little or no relationship between the
absolute amounts of the differences and the
book values.

Ratio Estimation is a classical variable sampling

technique that uses the ratio of audited
amounts to be recorded amounts in the sample
to estimate the total peso amount of the
population and an allowance for sampling risk.

Ratio estimation is more appropriate to use

when the difference is nearly proportional to
book values.

Conditions for using difference and ratio

estimation:

 Each population item must have a

recorded value

 Total population book value must be

known

 Expected difference between audited

and recorded book values must not be
too rare.

Probability-Proportional-to-Size Sampling (PPS)

/ Dollar-unit sampling

 This method uses attributes sampling

theory to express a conclusion in
peso amounts.

 Being a statistical method, it gives each

individual peso in the population an
equal chance of selection.

 Generally useful for verifying

“existence” assertions since, in this
method, the higher the amount, the
more likely that it will be selected.

 This method is primarily applicable in

testing account balances and
transactions for overstatement.