Statistical Sampling for Testing Control

Procedures
MULTIPLE CHOICE

1. Auditors who prefer statistical sampling to non-statistical

sampling may do so because statistical sampling helps the
auditor
a. Measure the sufficiency of the evidential matter
obtained.
b. Eliminate subjectivity in the evaluation of sampling
results.
c. Reduce the level of tolerable error to a relatively low
amount.
d. Minimize the failure to detect a material misstatement
due to non-sampling risk.

ANSWER: B

2. If all other factors in a sampling plan are held constant,

changing the measure of tolerable error to a smaller value
would cause the sample size to be:
a. Smaller.
b. Larger.
c. Unchanged.
d. Indeterminate.

ANSWER: B

3. Which of the following sampling plans would be designed to

estimate a numerical measurement of a population, such as a
dollar value?
a. Numerical sampling.
b. Discovery sampling.
c. Sampling for attributes.
d. Sampling for variables.

ANSWER: D

4. Based on a random sample, it is estimated that four percent,

plus or minus two percent, of a firm's invoices contain
errors. The plus or minus two percent is known as the
estimate's:
a. Precision.
b. Accuracy.

142
143 Chapter 9 Statistical Sampling for Testing Control Procedures

c. Confidence level.
d. Standard error.

ANSWER: A
5. In order to quantify the risk that sample evidence leads to
erroneous conclusions about the sampled population
a. Each item in the sampled population must have an equal
chance of being selected.
b. Each item in the sampled population must have a chance
of being selected proportional to its book value.
c. Each item in the sampled population must have an equal
or known probability of being selected.
d. The precise number of items in the population must be
known.

ANSWER: C

6. Which of the following statements is not true regarding

audit risk assessment?
a. The auditor studies the business and industry and
applies analytical procedures as a basis for assessing
inherent risk.
b. When control risk and inherent risk are high, the
auditor increases detection risk to maintain overall
audit risk at the desired level.
c. The auditor studies and evaluates internal control
policies and procedures for assessing control risk.
d. The auditor designs substantive audit procedures to
reduce detection risk to an acceptable level.

ANSWER: B

7. Which of the following factors does an auditor generally

need to consider in planning a particular audit sample
for a control test?
a. Number of items in the population.
b. Total dollar amount of the items to be sampled.
c. Estimated standard deviation of the population.
d. Tolerable error.

ANSWER: D

8. Random numbers can be used to select a sample only when each

item in the population:
a. Can be assigned to a specific stratum.
b. Is independent of outside influence.
c. Can be identified with a unique number.
144
Chapter 9 Statistical Sampling for Testing Control Procedures

d. Is expected to be within plus or minus three standard

deviations of the population mean.

ANSWER: C

9. The tolerable occurrence rate for a control test is

generally
a. Lower than the expected occurrence rate in the related
accounting records.
b. Higher than the expected occurrence rate in the related
accounting records.
c. Identical to the expected occurrence rate in the
related accounting records.
d. Unrelated to the expected occurrence rate in the
related accounting records.

ANSWER: B

10. An advantage of statistical over non-statistical sampling is

that statistical sampling:
a. Enables auditors to objectively measure the reliability
of their sample results.
b. Permits use of a smaller sample size than would be
necessary with non-statistical sampling.
c. Is compatible with a wider variety of sample selection
methods than is non-statistical sampling.
d. Allows auditors to inject their subjective judgment in
determining sample size and selection process in order
to audit items of greatest value and highest risk.

ANSWER: A

11. An auditor samples cash disbursement records for significant

errors of $5 or more. Upon finding one such error, these
records are scheduled for a complete review. This
conclusion is most likely based on a
a. Cluster sample.
b. Discovery sample.
c. Systematic sample.
d. Stratified sample.

ANSWER: B

12. Of the following statements, which one best differentiates

statistical sampling from nonstatistical sampling?
145 Chapter 9 Statistical Sampling for Testing Control Procedures

a. Statistical sampling is a mathematical approach to

inference, whereas nonstatistical sampling is a more
subjective approach.
b. Nonstatistical sampling has greater applicability to
large populations than does statistical sampling.
c. Nonstatistical sampling is more subjective, but
produces greater consistency in the application of
audit judgment.
d. Nonstatistical sampling has greater applicability to
populations that lend themselves to random selection.

ANSWER: A

13. In examining cash disbursements, an auditor plans to choose

a sample using systematic selection with a random start.
The primary advantage of such a systematic selection is that
population items
a. Which include errors will not be overlooked when the
auditor exercises compatible reciprocal
options.
b. May occur in a systematic pattern, thus making the
sample more representative.
c. May occur more than once in a sample.
d. Do not have to be prenumbered in order for the auditor
to use the technique.

ANSWER: D

14. When planning an attribute sampling application, the

difference between the expected occurrence rate and the
tolerable occurrence rate is the planned
a. Precision.
b. Reliability.
c. Dispersion.
d. Skewness.

ANSWER: A

15. If certain forms are not consecutively numbered

a. Selection of a random sample probably is not possible.
b. Systematic sampling may be appropriate.
c. Stratified sampling should be used.
d. Random number tables cannot be used.

ANSWER: B
146
Chapter 9 Statistical Sampling for Testing Control Procedures

16. In an attribute sampling application, holding other factors

constant, sample size will increase as which of the
following becomes smaller?
a. Reliability level.
b. Population.
c. Planned precision.
d. Expected rate of occurrence.

ANSWER: C

17. Which of the following sampling methods is most useful to

auditors when testing for internal control effectiveness?
a. Stratified random sampling.
b. Attribute sampling.
c. Variables sampling.
d. Unrestricted random sampling with replacement.

ANSWER: B

18. Attribute sampling, as applied to control testing, can

assist the auditor in several ways. Which of the following
tasks is not enhanced by sampling?
a. Determining the number of documents to examine in
testing for a specific attribute.
b. Selecting the documents to be tested.
c. Examining the documents.
d. Evaluating the sample results.

ANSWER: C

19. Which of the following is an element of sampling risk?

a. Choosing an audit procedure that is inconsistent with
the audit objective.
b. Choosing a sample size that is too small to achieve the
sampling objective.
c. Failing to detect an error on a document that has been
inspected by the auditor.
d. Failing to perform audit procedures that are required
by the sampling plan.

ANSWER: B

20. Which of the following would not be an attribute of interest

to an auditor performing control tests?
a. Do selling prices agree with published price lists?
b. Do purchased parts meet established quality standards?
c. Are proper labor rates being used to compute payroll?
147 Chapter 9 Statistical Sampling for Testing Control Procedures

d. Are account distributions (debits and credits) correct?

ANSWER: B

21. The precision limit for control testing necessary to

justify lowering the assessed control risk level
depends primarily on which of the following?
a. The cause of the errors.
b. The materiality of the attribute(s) to be tested.
c. The amount of any substantive errors.
d. The limit used in audits of similar clients.

ANSWER: B

22. Which of the following statements concerning sample size is

true?
a. An increase in the tolerable occurrence rate, other
factors remaining unchanged, increases sample size.
b. The higher the expected occurrence rate, other factors
remaining unchanged, the larger will be the sample
size.
c. The more critical the attribute being tested, the
higher will be the tolerable occurrence rate set by the
auditor, and the larger will be the sample size.
d. The lower the acceptable risk of underassessment of
control risk, the smaller will be the sample size.

ANSWER: B

23. In the examination of the financial statements of Delta

Company, the auditor determines that in performing a
test of internal control effectiveness, the rate of error in
the sample does not support the auditor's preconceived
notion of a tolerable occurrence rate when, in fact, the
actual error rate in the population does meet the auditor's
notion of effectiveness. This situation illustrates
the risk of
a. Underassessment of control risk.
b. Overassessment of control risk.
c. Incorrect rejection.
d. Incorrect acceptance.

ANSWER: B

24. If all other factors in a sampling plan are held constant,

changing the risk of underassessment from five percent to
148
Chapter 9 Statistical Sampling for Testing Control Procedures

three percent would cause the sample size to be:

a. Smaller.
b. Larger.
c. Unchanged.
d. Indeterminate.

ANSWER: B

25. An auditor performs a test to determine whether all

merchandise for which the client was billed was
received. The population for this test consists of all
a. Merchandise received.
b. Vendors' invoices.
c. Canceled checks.
d. Receiving reports.

ANSWER: B

26. Although mathematically based, statistical sampling does not

replace audit judgment. In utilizing statistical sampling
techniques, the auditor must apply judgment in all but which
of the following tasks?
a. Selecting a tolerable rate of error.
b. Determining an acceptable risk of underassessing
control risk.
c. Calculating the actual error rate.
d. Assessing the materiality of control weaknesses.

ANSWER: C

27. To determine sample size in an attribute sampling

application, what must be specified?
a. Population mean, expected error rate, precision.
b. Precision, reliability, standard deviation.
c. Precision, reliability, expected occurrence rate.
d. Population mean, standard deviation, precision.

ANSWER: C

28. Which of the following statements regarding statistical

sampling in auditing is true?
a. Inasmuch as audits are test-based, generally accepted
auditing standards require the use of statistical
sampling methods whenever the auditor decides to
examine only part of a population.
b. Although statistical sampling may be applied to control
testing, it is required for substantive testing
149 Chapter 9 Statistical Sampling for Testing Control Procedures

purposes.
c. Sampling methods are used by auditors in both control
testing and substantive testing.
d. Statistical sampling methods are more appropriate for
control testing when the auditor elects to reprocess
transactions, than when controls are tested by means of
document examination.

ANSWER: C

29. The application of statistical sampling techniques is least

related to which of the following generally accepted
auditing standards?
a. The work is to be adequately planned and assistants, if
any, are to be properly supervised.
b. In all matters relating to the assignment, an independ-
ence in mental attitude is to be maintained by the
auditor or auditors.
c. A sufficient understanding of the internal control
system is to be obtained to plan the audit and to
determine the nature, timing, and extent of tests
to be performed.
d. Sufficient competent evidential matter is to be
obtained through inspection, observation, inquiries,
and confirmations to afford a reasonable basis for
an opinion regarding the financial statements
under examination.

ANSWER: B

30. The major weakness of nonstatistical (judgmental) sampling

is it
a. Usually requires larger sample sizes than statistical
sampling.
b. Does not allow sampling risk to be objectively
measured.
c. Frequently results in samples that are not
representative of the population.
d. Gives less accurate point estimates of parameters than
statistical sampling.

ANSWER: B

31. An auditor is testing credit authorization procedures by

examining sales invoices for credit approval by the credit
department. The procedures will be considered to be working
adequately if 96% of all sales invoices either indicate
150
Chapter 9 Statistical Sampling for Testing Control Procedures

approval or are cash sales. The auditor selects a

random sample of 100 invoices. In this situation, which
of the following outcomes illustrates underassessment?
a. The auditor finds five deviations and concludes that
procedures work inadequately. The actual population
deviation rate is 2%.
b. The auditor finds no deviations and concludes that
procedures work adequately. The true population
deviation rate is 5%.
c. The auditor finds no deviations and concludes that the
procedures work adequately. The true population
deviation rate is 2%.
d. The auditor finds five deviations and concludes that
procedures work inadequately. The true population
deviation rate is 6%.

ANSWER: B

32. If an auditor, planning to use statistical sampling, is

concerned with the number of a client's sales invoices
that contain mathematical errors, the auditor would most
likely utilize
a. Random sampling with replacement.
b. Sampling for attributes.
c. Sampling for variables.
d. Stratified random sampling.

ANSWER: B

33. An advantage of using statistical over non-statistical

sampling methods in tests of controls is that the
statistical methods
a. Afford greater assurance than a non-statistical sample
of equal size.
b. Provide an objective basis for quantitatively
evaluating sample risks.
c. Can more easily convert the sample into a dual-purpose
test useful for substantive testing.
d. Eliminate the need to use judgment in determining
appropriate sample sizes.

ANSWER: B

34. Several risks are inherent in the evaluation of audit

evidence which has been obtained through the use of
statistical sampling. Which of the following
151 Chapter 9 Statistical Sampling for Testing Control Procedures

risks is an example of the risk of underassessment of control

risk?
a. Failure to properly define the population to be
sampled.
b. Failure to draw a random sample from the population.
c. Failure to accept the statistical hypothesis that
internal control is unreliable when, in fact, it is.
d. Failure to accept the statistical hypothesis that a
book value is not materially misstated when the true
book value is not materially misstated.

ANSWER: C

35. As a result of tests of controls, an auditor underassessed

control risk and decreased substantive testing. This
underassessment occurred because the true occurrence rate in
the population was
a. Less than the risk of underassessment in the auditor's
sample.
b. Less than the occurrence rate in the auditor's sample.
c. More than the risk of underassessment in the auditor's
sample.
d. More than the occurrence rate in the auditor's sample.

ANSWER: D

36. If all other factors specified in an attribute sampling plan

remain constant, changing the specified precision from 6% to
10%, and changing the specified reliability from 97% to 93%
would cause the required sample size to
a. Increase.
b. Remain the same.
c. Decrease.
d. Change by 4%.

ANSWER: C

37. Which of the following factors is generally not considered

in determining the sample size for a test of controls?
a. Population size.
b. Tolerable rate.
c. Risk of underassessment.
d. Expected population occurrence rate.

ANSWER: A
152
Chapter 9 Statistical Sampling for Testing Control Procedures

38. The expected population occurrence rate of client billing

errors is 3%. The auditor has established a tolerable
rate of 5%. In the review of client invoices the auditor
should use
a. Stratified sampling.
b. Variable sampling.
c. Discovery sampling.
d. Attribute sampling.

ANSWER: D

39. An auditor desired to test credit approval on 10,000 sales

invoices processed during the year. The auditor designed a
statistical sample that would provide a 1% risk of assessing
control risk too low (99% confidence) that not more than 7%
of the sales invoices lacked approval. The auditor estimated
from previous experience that about 2 1/2% of the sales
invoices lacked approval. A sample of 200 invoices was
examined and 7 of them were lacking approval. The
auditor then determined the upper occurrence limit to be
8%. In the evaluation of this sample, the auditor decided to
increase the level of the preliminary assessment of control
risk because the
a. Tolerable rate (7%) was less than the achieved upper
occurrence limit (8%).
b. Expected occurrence rate (7%) was more than the
percentage of errors in the sample (3 1/2%).
c. Achieved upper occurrence limit (8%) was more than the
percentage of errors in the sample (3 1/2%).
d. Expected occurrence rate (2 1/2%) was less than the
tolerable rate (7%).

ANSWER: A

40. To evaluate the sample results for an attributes sampling

application, a measure of precision is required. Which of
the following statements is true regarding precision?
a. Precision is subtracted from the tolerable rate of
error to determine the acceptability of sample
results. b. Precision is added to the tolerable rate
of error to determine the acceptability of sample
results. c. Precision is subtracted from the
sample occurrence rate and the result is compared to the
tolerable rate of error.
d. Precision is added to the sample occurrence rate and
the result is compared to the tolerable rate of error.
153 Chapter 9 Statistical Sampling for Testing Control Procedures

ANSWER: D

41. An auditor who uses statistical sampling for attributes in

testing internal controls should increase the assessed level
of control risk when the
a. Sample occurrence rate is less than the expected
occurrence rate used in planning the sample.
b. Tolerable rate less the allowance for sampling risk
exceeds the sample occurrence rate.
c. Sample occurrence rate plus the allowance for sampling
risk exceeds the tolerable rate.
d. Sample occurrence rate plus the allowance for sampling
risk equals the tolerable rate.

ANSWER: C

42. Which of the following statements is correct concerning

statistical sampling for control testing?
a. The population size has little or no effect on
determining sample size except for very small
populations.
b. The expected population occurrence rate has little or
no effect on determining sample size except for
very small populations.
c. As the population size doubles, the sample size also
should double.
d. For a given tolerable rate, a larger sample size should
be selected as the expected population deviation rate
decreases.

ANSWER: A

43. An auditor wishes to determine if the error rate on travel

reimbursement claims is within the five-percent tolerance
level set by management. What sampling plan should the
auditor use?
a. Variables sampling.
b. Attribute sampling.
c. Judgment sampling.
d. PPS sampling.

ANSWER: B

44. The auditor samples cash disbursements for minor errors of

$5 or less. The parameter to be estimated is the error
rate, which might be as high as 10%. The auditor is
most likely to use
154
Chapter 9 Statistical Sampling for Testing Control Procedures

a. Block sampling.
b. Variables sampling.
c. Attributes sampling.
d. Discovery sampling.

ANSWER: C

45. An important difference between a statistical sample and a

non-statistical (judgmental) sample is that with a
statistical sample:
a. No judgment is required, everything is by formula.
b. A smaller sample size can be used.
c. More accurate results are obtained.
d. Population estimates with measurable reliability can be
made.

ANSWER: D

COMPLETION:

47. Statistical sampling for attributes is most applicable in

situations where a visible in the form
of exists.

ANSWER: AUDIT TRAIL, DOCUMENTATION

48. As contrasted with statistical sampling, nonstatistical

sampling is a more approach to inference.

ANSWER: SUBJECTIVE

49. estimates the frequency of events,

whereas is an estimate of amount.

ANSWER: ATTRIBUTE SAMPLING, VARIABLES SAMPLING

50. The is the anticipated

error rate, whereas the is
the maximum rate of error acceptable to the auditor, while
still warranting a lowering of assessed control risk below
the maximum level.

ANSWER: EXPECTED OCCURRENCE RATE, TOLERABLE OCCURRENCE RATE

155 Chapter 9 Statistical Sampling for Testing Control Procedures

51. The the precision range, the larger the sample

size.

ANSWER: NARROWER

52. The more critical an attribute to those control policies and

procedures relevant to an audit, the the tolerable
occurrence rate.

ANSWER: LOWER

53. Another term for "risk of underassessment" is ____________

` risk.

ANSWER: BETA

54. Random selection requires that items to be included in the

sample must be drawn on a basis.

ANSWER: PROBABILITY

55. When documents are not prenumbered, an acceptable

alternative to the use of random number tables or computer-
generated numbers is .

ANSWER: SYSTEMATIC SAMPLING

56. Under conditions of detection risk, a minimal

quantity of audit evidence is needed.

ANSWER: HIGH

MATCHING:

57. Indicate by letter the term matching the definition given.

A. Attribute sampling
B. Alpha risk
C. Discovery sampling
D. Beta risk
E. Expected occurrence rate
F. Precision
G. Discovery sampling
H. Sampling risk
I. Systematic sampling
J. Tolerable occurrence rate
K. Upper occurrence limit
156
Chapter 9 Statistical Sampling for Testing Control Procedures

L. Variables sampling

____1. An estimate of frequency of events.

____2. An orderly approach to locating a particular event.

____3. The anticipated error rate.

____4. Risk of underassessment.

____5. Maximum rate of error acceptable to the auditor.

____6. Risk that the auditor’s conclusions about a population

will be incorrect.

____7. A random selection method that involves choosing every

nth item in the population until the requisite sample
size has been reached.

____8. The calculated maximum error rate based on the results of

sampling.

____9. Risk of overassessment.

____10. Range within the true answer most likely falls.

SOLUTION:

1. A
2. C
3. E
4. D
5. J
6. H
7. I
8. K
9. B
10. F

PROBLEM/ESSAY:

58. Margaret Hopkins, CPA, wishes to test the effectiveness of

internal controls over the customer billing function of
Jack’s Wholesale Hardware, an audit client. Hopkins has
identified the following attributes to be tested. For each
attribute, indicate the appropriate sampling unit and
carefully define what constitutes an error.
157 Chapter 9 Statistical Sampling for Testing Control Procedures

1. Does the customer order agree with the invoice as to

type and quantity of goods?
2. Does the shipping order agree with the invoice as to
type and quantity of goods?
3. Has customer credit been approved?
4. Does the invoice amount cause the customer to exceed
the prepared credit limit?
5. Have goods been billed but not shipped?
6. Have goods been shipped but not billed?
7. Do invoiced prices agree with official price lists?
8. Do credit and discount terms appearing on face of
customer invoice accord with company policy?
9. Are proper accounts debited and credited?

SOLUTION:

1. SAMPLING UNIT: Sales invoice

ERROR: Lack of agreement between invoice and order
as to:
Type of goods
Quantity of goods
Price of goods
Credit and discount terms

2. SAMPLING UNIT: Sales invoice

ERROR: Lack of agreement between invoice and
shipping order as to:
Type of goods
Quantity of goods

3. SAMPLING UNIT: Sales invoice

ERROR: Lack of evidence of credit approval on face
of invoice

4. SAMPLING UNIT: Sales invoice

ERROR: Invoice amount plus previous customer
balance exceeds prepared credit limit

5. SAMPLING UNIT: Sales invoice

ERROR: Shipping order and/or bill of lading not
attached to customer invoice

6. SAMPLING UNIT: Bill of lading

ERROR: Matching invoice not located

7. SAMPLING UNIT: Sales invoice

158
Chapter 9 Statistical Sampling for Testing Control Procedures

ERROR: Price not in agreement with master price list

and no evidence of proper approval of
billed price

8. SAMPLING UNIT: Sales invoice

ERROR: Credit and/or discount terms at variance with
company policy, and no evidence of proper
approval of quoted credit and discount
terms.

9. SAMPLING UNIT: Sales invoice

ERROR: Incorrect debits or credits appearing on face
of invoice and not corrected before posting
accounts.

59. Sampling can assist the auditor in various ways during the
examination of documentary evidence for tests of
internal
controls. Using a vouchering and vendor billing application
as an example, assume that you are interested in evaluating
internal controls relative to ordering goods, receiving
goods, and processing vendor invoices. Discuss how
attribute sampling might assist you in the following
phases
of your tests of controls:
a. Determining the number of vouchers to test;
b. Selecting the vouchers for testing;
c. Evaluating the results of the voucher test;
d. Assessing audit risk associated with the
expenditure cycle.

SOLUTION:

a. In determining the number of vouchers to test, the

auditor will need to set a tolerable occurrence rate and an
acceptable risk of underassessment. These parameters,
along with an estimated occurrence rate may then be used as
inputs to determining sample size.
b. If vouchers are prenumbered, the auditor can elect to use
a random selection approach to selecting vouchers test
for predefined attributes.
c. Having examined the sample for the attributes of
interest, the auditor can calculate the observed error
rate and the upper occurrence limit.
d. The upper occurrence limit can be compared with the
tolerable occurrence rate to support raising or lowering
the assessed level of control risk.
159 Chapter 9 Statistical Sampling for Testing Control Procedures

60. George Jones, CPA, wishes to test payroll controls for

Hopkins Manufacturing, an audit client. He has identified
the following attributes of interest for testing purposes:
1. Correctness of pay rates and hours used in
preparing the payroll summary;
2. Genuineness of listed employees;
3. Correctness of gross pay, withholdings, and net
pay;
4. Correctness of debits and credits.
A pilot test of forty payroll transactions revealed the
following errors: One transaction revealed an
incorrect
pay rate; two transactions resulted in incorrect
calculations of either gross pay or withholdings; and two
transactions resulted in erroneous debits or credits. All
listed employees were found to be working for the company
during the pay period being tested. Jones believes that
four or more errors for each of the four attributes is
significant.

Required:
a. Based on the pilot sample, calculate the expected
occurrence rate and the tolerable occurrence rate for
each attribute.
b. In determining sample size, Jones can select an
expected occurrence rate ranging from the low for
attribute number 2 to the high for attributes 3 and 4.
What impact will a low rate relative to a high rate
have on the calculated sample size?
c. What other factors must Jones consider before
determining the sample size?
c. Assuming that a sample of 100 payroll summary entries
produced the following calculated upper occurrence
limits, what impact will these limits have on
Jones substantive audit program?

Attribute 1: 7.0%
Attribute 2: 4.0%
Attribute 3: 12.0%
Attribute 4: 6.0%

SOLUTION:

a. The tolerable occurrence rate is 10% (4/40) for each

160
Chapter 9 Statistical Sampling for Testing Control Procedures

of the attributes. The expected occurrence rates are:

Attribute No. 1: 2.5%
Attribute No. 2: 0.0%
Attribute No. 3: 5.0%
Attribute No. 4: 5.0%
b. The higher the expected occurrence rate, the larger
will be the sample size.
c. Jones must determine the acceptable risk of
underassessing control risk.
d. Only attribute 3 produces an unacceptable upper
occurrence limit, exceeding the tolerable rate of 10%.
Given incorrect calculation of gross pay, withholdings,
and net pay, Jones should plan to focus more heavily
on inventory costing, inasmuch as inventory includes
both direct and indirect labor components. He should
also concentrate on payroll cost included in general
and administrative expense, as well as on accrued
payroll and payroll taxes payable at year end.
Incorrect calculation of gross pay and withholdings
will also impact the calculation of employer’s payroll
tax expense. Therefore, Jones must plan to audit these
accounts more intensively than under conditions of
strong internal control.