Cia Review: Part 1 Study Unit 6: Sampling and Statistical Quality Control

Table of Contents
6.1 Statistical Concepts

CIA REVIEW: PART 1 6.2 Sampling Concepts
6.3 Attribute Sampling
Study Unit 6 6.4 Variables Sampling
Sampling and Statistical 6.5 Statistical Quality Control
Quality Control
Copyright © 2017 Gleim Publications, Inc. All rights reserved. Duplication prohibited. Reward for information exposing violators. Contact copyright@gleim.com. 1 Copyright © 2017 Gleim Publications, Inc. All rights reserved. Duplication prohibited. Reward for information exposing violators. Contact copyright@gleim.com. 2
CIA 1 SU 6 CIA 1 SU 6
Populations and Samples

Sampling involves selecting representative items from a
population (an entire group of items), examining those
selected items, and drawing a conclusion about the
population based on the results derived from the
examination of the selected items.
Statistical Concepts
6.1
Population Distributions Measures of Central Tendency
For audit purposes, each item in a population is associated
with a variable of interest to the auditor. curve are quantified through its measures of central
o Discrete variables are tested using attribute sampling. tendency.
o Continuous variables are tested using variables o The mean is the arithmetic average of a set of numbers.
sampling. o The median is the middle value if data are arranged in
An important characteristic of a population is the numerical order.

distribution of the values of the variable of interest. o The mode is the most frequently occurring value.
Example of Mean,
Median, and Mode Normal Distribution
In a normal distribution, the mean, median, and mode are
the same, and the tails are identical.
Asymmetrical Asymmetrical
Frequency Distribution Frequency Distribution
In some asymmetrical frequency distributions, the mean is In some asymmetrical frequency distributions, the median
greater than the mode. The right tail is longer, and the is greater than the mean. The left tail is longer, and the
distribution is positively skewed (to the right). distribution is negatively skewed (to the left).
Standard Deviation and Confidence Example of a

Level for Normal Distributions Normal Distribution
A certain species of pine tree has an average adult height of 20 feet, with
items are spread about the mean (dispersion). It is each standard deviation representing 1 foot. The conclusion from the
measured by the standard deviation. distribution below is that 68% of all trees of this species will reach a height
between 19 and 21 feet (1 standard deviation), 95.5% will be between
Normal distributions may have the following fixed 18 and 22 feet (2 standard deviations), and 99.7% will be between 17 and
relationships between the area under the curve and the 23 feet (3 standard deviations).
distance from the mean:
Confidence Level and
Confidence Interval Example
The confidence (reliability) level is the percentage of times An auditor took a random sample of sales authorizations.
that a sample is expected to be representative of the Based on the sample, the sales department authorized a
population; i.e., a confidence level of 95% should result in sale after checking the credit score of the customer 88% of
representative samples 95% of the time. the time.
o If the confidence interval (precision) is 6%, the auditor
The confidence interval (precision), also known as the can be confident that between 82% (88% 6%) and 94%
allowance for sampling risk, based on a specified confidence
level, is the range around a sample value that is expected to after checking credit scores.
contain the true population value. o
the sample. If the specified confidence level is 95% and
the precision is 6%, the auditor can be 95% confident
authorized after checking credit scores is between 82%
and 94%.
Pilot Sampling
and Standard Error Multiple-Choice Question
The auditor can estimate the standard deviation of a A 90% confidence interval for the mean of a population based on the information
in a sample always implies that there is a 90% chance that the
population using a pilot sample.
Standard error of the mean is the standard deviation of the A. Estimate is equal to the true population mean.
distribution of sample means.
B. True population mean is no larger than the largest endpoint of the interval.
Coefficient of variability measures the relative variability C. Standard deviation will not be any greater than 10% of the population mean.
within the data and is calculated by dividing the standard
D. True population mean lies within the specified confidence interval.
deviation of the sample by the mean.
Multiple-Choice Answer Multiple-Choice Question
A 90% confidence interval for the mean of a population based on the information One standard deviation corresponds to what approximate percentage of the area
in a sample always implies that there is a 90% chance that the under a bell curve?
A. Estimate is equal to the true population mean. A. 34%

B. True population mean is no larger than the largest endpoint of the interval. B. 68%
C. Standard deviation will not be any greater than 10% of the population mean. C. 90%
D. True population mean lies within the specified confidence interval. D. 95.5%
The confidence level, e.g., 90%, is specified by the auditor. A confidence
interval based on the specified confidence level, also called precision, is
the range around a sample value that is expected to contain the true
population value. In this situation, if the population is normally
distributed and repeated simple random samples are taken, the
probability is that 90% of the confidence intervals constructed around the
sample results will contain the population value.
Multiple-Choice Answer
One standard deviation corresponds to what approximate percentage of the area
under a bell curve?
A. 34%
B. 68%
C. 90%
D. 95.5%
Sampling Concepts
A person selecting an item at random from a normally distributed
population (one with a bell curve) can be 68% confident that the
value of the item is within 1.0 standard deviation of the mean. Thus,
6.2
that area under the curve corresponds to the 68% confidence level.
Nonstatistical
(Judgmental) Sampling Statistical Sampling
Statistical sampling provides an objective method of determining
determine the sample size and sample selection. sample size and selecting the items to be examined.
Advantages Advantages
o The process can be less expensive and less time consuming. o It provides a quantitative measure of sampling risk,
o The auditor has greater discretion to use his or her judgment confidence level, and precision.
and expertise. o It provides a quantitative expression of sample results.
Disadvantages o It helps the auditor design an efficient sample.
o It does not provide a quantitative measure of sampling risk. Disadvantages

o It does not provide a quantitative expression of sample o It can be more expensive and time consuming than
results. nonstatistical sampling.
o If the auditor is not proficient, the sample may not be o It requires special statistical knowledge and training.
effective. o It requires statistical software.
Nonsampling vs.
Sampling Risk Selecting the Sample Approach
Nonsampling risk is audit risk not related to sampling. In a random sample, every item in the population has an
o equal and nonzero chance of being selected.
error in a sample. An interval (systematic) sampling plan assumes that items
Sampling risk is the risk that a sample is not representative are arranged randomly in the population. If the auditor
of the population. discovers this is not true, a random selection method should
o An unrepresentative sample may result in an incorrect
be used.
conclusion. Block (cluster) sampling randomly selects groups of items
as the sampling units rather than individual items. An
example is the inclusion in the sample of all cash payments
for May and September.
Multiple-Choice Question Multiple-Choice Answer
If an internal auditor is sampling to test compliance with a particular company If an internal auditor is sampling to test compliance with a particular company
policy, which of the following factors should not affect the allowable level of policy, which of the following factors should not affect the allowable level of
sampling risk? sampling risk?
A. The experience and knowledge of the auditor. A. The experience and knowledge of the auditor.
B. The adverse consequences of noncompliance. B. The adverse consequences of noncompliance.
C. The acceptable level of risk of making an incorrect audit conclusion. C. The acceptable level of risk of making an incorrect audit conclusion.
D. The cost of performing auditing procedures on sample selections. D. The cost of performing auditing procedures on sample selections.
Sampling risk is the possibility that engagement conclusions

based on a sample may differ from those reached if the test
were applied to all items in the population. The experience and
knowledge of the auditor are elements of nonsampling risk.
Uses
In attribute sampling, each item in the population has an
attribute of interest to the auditor, e.g., evidence of proper
authorization.
Attribute sampling is appropriate for tests of controls, i.e.,
when two outcomes are possible (compliance or
noncompliance).
Attribute Sampling
6.3
Factors Affecting
Determining the Sample Size Attribute Sample Size
The sample size for an attribute test depends on the
following four factors:
o Confidence level is the percentage of times that a
sample is expected to be representative of the
population.
o Population size is the sum of the items to be considered
for testing.
o Expected deviation rate is an estimate of the deviation
rate in the current population.
o Tolerable deviation rate is the highest allowable
percentage of the population that can be in error and
still allow the auditor to rely on the tested control.
Two Other Attribute

Evaluation of Sample Results Sampling Methods
The sample deviation rate is the number of deviations Discovery sampling is appropriate when even a single
divided by the sample size. deviation (noncompliance) is critical.
The achieved upper deviation limit (UDL) is based on the The objective of stop-or-go sampling, also called sequential
sample size and the number of deviations discovered. sampling, is to reduce the sample size when the auditor
Auditors use standard tables to calculate the UDL. believes the deviation rate in the population is low.
In selecting a sample of items for attributes testing, an auditor must consider the In selecting a sample of items for attributes testing, an auditor must consider the
confidence level factor, the desired precision, and the confidence level factor, the desired precision, and the
A. Recorded monetary amount of the population. A. Recorded monetary amount of the population.
B. Sampling interval. B. Sampling interval.
C. Expected occurrence rate. C. Expected occurrence rate.
D. Standard deviation in the population. D. Standard deviation in the population.
The expected occurrence rate, also called the expected

deviation rate, is one of the three necessary factors in
determining sample size for an attribute test.
Uses
Variables sampling provides information about whether a
stated amount (e.g., the balance of accounts receivable) is
materially misstated.
In variables sampling, both the upper and lower limits are
relevant.
Auditors may employ the following variables sampling
Variables Sampling techniques:
o Unstratified mean-per-unit
o Stratified mean-per-unit
6.4 o Difference estimation
o Ratio estimation
o Monetary unit sampling
Factors Affecting
Determining the Sample Size Variables Sample Size
The sample size for a variable test depends on the following
four factors:
o Confidence level
o Population size
o Tolerable misstatement
o Standard deviation
Primary Methods of Primary Methods of

Variables Sampling Variables Sampling
Mean-per-unit (MPU) estimation averages the audited Ratio estimation estimates the population misstatement by
amounts of the sample items. It multiplies the average by multiplying the recorded amount of the population by the
the number of items in the population to estimate the ratio of the total audited amount of the sample items to
population amount. An achieved precision at the desired their total recorded amount.
level of confidence is then calculated. Monetary-unit sampling (MUS), also known as probability-
Difference estimation estimates the misstatement of an proportional-to-size (PPS) sampling, uses a monetary unit as
amount by calculating the difference between the observed the sampling unit and applies attribute sampling methods to
and recorded amounts for items in the sample. reach a conclusion about the probability of overstating an
account balance.
Example
An auditor examines a sample of 150 accounts receivable with a total recorded amount of US $172,500. The total population of
Multiple-Choice Question
3,400 accounts receivable has a total recorded amount of US $3,500,000. Based on the audit, the total amount of the 150
sampled accounts is US $168,000.
MPU Estimation
The average amount per sampled item is US $1,120 ($168,000 ÷ 150).
If all other factors in a sampling plan are held constant, changing the measure of
The estimated correct balance of the population (accounts receivable) is US $3,808,000 ($1,120 mean per unit value × 3,400 tolerable misstatement to a smaller value will cause the sample size to be
number of items in the population).
Difference Estimation
The difference between the audited and recorded amounts of items in the sample is US $4,500 ($172,500 $168,000). A. Smaller.
The mean difference is US $30 ($4,500 ÷ 150 number sample items).
B. Larger.
The estimated total population error is determined by multiplying the mean by the number of items in the population. It
equals US $102,000 (3,400 × $30). C. Unchanged.
The estimated correct balance of the population (accounts receivable) is US $3,398,000 ($3,500,000 recorded amount of the
population $102,000 estimated error). D. Indeterminate.
Ratio Estimation
The ratio of the total audited amount of the sample items to their total recorded amount is 0.974 (US $168,000 audited
amount ÷ $172,500 recorded amount).
The estimated correct balance of the population (accounts receivable) is US $3,409,000 ($3,500,000 recorded amount of the
population × 0.974 ratio).
NOTE: An achieved precision at the desired level of confidence is then calculated. For example, assume the sample of 150
accounts with a total amount of US $168,000 was based on precision of ±3% and a confidence level of 95%. Using ratio
estimation, the precision interval equals ±US $102,270 ($3,409,000 × 3%). The auditor can conclude that the probability is only 5%
that the true balance lies outside the range of US $3,306,730 to US $3,511,270.
Multiple-Choice Answer
If all other factors in a sampling plan are held constant, changing the measure of
tolerable misstatement to a smaller value will cause the sample size to be
A. Smaller.
B. Larger.
C. Unchanged.
D. Indeterminate.
The size of the precision interval in a variables test is based upon the
Statistical Quality Control
tolerable misstatement that is determined by materiality judgments. As
this value decreases, for example, because of a decrease in tolerable 6.5
misstatement, the size of the required sample increases accordingly,
and vice versa. Hence, tolerable misstatement (precision) and sample
size are inversely related.
Uses Acceptance Sampling
Statistical quality control determines whether a shipment or This method determines the probability that the rate of
production run of units lies within acceptable limits. defective items in a batch is less than a specified level.
Items are either good or bad, i.e., inside or outside of EXAMPLE:
control limits. Assume a sample is taken from a population of 500.
It is also used to determine whether production processes According to standard acceptance sampling tables, if the
are out of control. sample consists of 25 items and not one is defective, the
probability is 93% that the population deviation rate is less
than 10%. If 60 items are examined and no defects are
found, the probability is 99% that the deviation rate is less
than 10%. If two defects in 60 units are observed, the
probability is 96% that the deviation rate is less than 10%.
Statistical Control Charts Example

Statistical control charts are graphic aids for monitoring the The chart below depicts 2 weeks of production by a manufacturer who produces a single
precision part each day. To be salable, the part can vary from the standard by no more than
status of any process subject to acceptable or unacceptable ± 0.1 millimeter.
variations during repeated operations.
A control chart consists of three lines plotted on a horizontal
time scale.
o The center line represents the overall mean or average
range for the process being controlled.
o The other two lines are the upper control limit (UCL) and
the lower control limit (LCL).
The part produced on the 20th had to be scrapped, and changes were made to the
Another advantage of the chart is that it makes trend and
cycles visible.
Cost-Benefit Analysis Pareto Diagrams
An analysis using expected value provides a more objective A Pareto diagram is a bar chart that assists managers in
basis for setting control limits. what is commonly called 80:20 analysis.
The limits of controls should be set so that the cost of an o The 80:20 rule states that 80% of all effects are the
investigation is less than or equal to the benefits derived. result of only 20% of all causes.
The benefit of an investigation is the avoidance of the costs The independent variable, plotted on the x axis, is the factor
of continuing to operate an out-of-control process. selected by the manager as the area of interest:
department, time period, geographical location, etc.
The frequency of occurrence of the defect (dependent
variable) is plotted on the y axis.
Example Histograms
A histogram displays a continuous frequency distribution of
A chief administrative the independent variable.
officer uses a Pareto This histogram shows the CAO the amount of travel
diagram to view which reimbursement delayed by a typical returned travel voucher.
departments are
generating the most
travel vouchers that have
been rejected because of
incomplete
documentation.
Fishbone Diagrams Fishbone Diagram
A fishbone (Ishikawa) diagram (also called a cause-and-
effect diagram) is a total quality management process
improvement technique.
This format organizes the analysis of causation and helps to
identify possible interactions among causes.
o The head of the skeleton contains the statement of the
problem.
o The principle classifications of causes are represented by
lines (bones) drawn diagonally from the heavy horizontal
line (the spine).
o Smaller horizontal lines are added in their order of
probability in each classification.

An organization has collected data on the complaints made by personal computer users and has An organization has collected data on the complaints made by personal computer users and has
categorized the complaints. The chart displays the categorized the complaints. The chart displays the
A. Arithmetic mean of each computer complaint. A. Arithmetic mean of each computer complaint.
B. Relative frequency of each computer complaint. B. Relative frequency of each computer complaint.
C. Median of each computer complaint. C. Median of each computer complaint.
D. Absolute frequency of each computer complaint. D. Absolute frequency of each computer complaint.
This Pareto diagram depicts the frequencies of complaints in absolute terms. It

displays the actual number of each type of complaint. The chart does not display
arithmetic means, relative frequencies, or medians of each type of complaint.

Cia Review: Part 1 Study Unit 6: Sampling and Statistical Quality Control

Uploaded by

Copyright:

Available Formats

Cia Review: Part 1 Study Unit 6: Sampling and Statistical Quality Control

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cia Review: Part 1 Study Unit 6: Sampling and Statistical Quality Control

Uploaded by

Copyright:

Available Formats

Table of Contents

6.1 Statistical Concepts

Sampling and Statistical 6.5 Statistical Quality Control

Populations and Samples

An important characteristic of a population is the numerical order.

Standard Deviation and Confidence Example of a

A. Estimate is equal to the true population mean. A. 34%

Disadvantages o It helps the auditor design an efficient sample.

o It does not provide a quantitative measure of sampling risk. Disadvantages

Sampling risk is the possibility that engagement conclusions

Two Other Attribute

The expected occurrence rate, also called the expected

Primary Methods of Primary Methods of

Statistical Control Charts Example

Multiple-Choice Question Multiple-Choice Answer

This Pareto diagram depicts the frequencies of complaints in absolute terms. It

You might also like