Statistical sampling is NOT based on which of the following audit sample techniques?

Statistical sampling is NOT based on which of the following audit sample techniques?
A . Haphazard Sampling
B . Random Sampling
C . Cell Sampling
D . Fixed interval sampling

View Answer

Answer: A

Explanation:

The NOT keyword is used in the question. You need find out an option which is NOT an example of statistical sampling. Statistical sampling is NOT based on Haphazard sampling.

For your exam you should know the information below

Audit samples are selected for the purpose of collecting representative evidence to be subjected to either compliance testing or substantive testing. The auditor should consider a selection technique that will provide the most relevant evidence supported by appropriate analytical procedures.

Two basic types of audit samples can be designed by the auditor to fulfill their requirements:

statistical and no statistical. Below Figure shows the various audit samples, as well as their testing methods. Care is given to the selection process in order to avoid drawing the wrong conclusion from the wrong sample. This is referred to as a sampling risk. Let’s look at each of these samples more closely.

Statistical Sampling

Statistical sampling uses mathematical techniques that result in an outcome that is mathematically quantifiable. Statistical samples are usually presented as a percentage. The purpose of statistical sampling is to gain an objective representation. Samples are selected by an objective mathematical process. The auditor should be aware that if the client has strong internal controls, the sample sizes may be smaller because the odds of fraud or failure will be lower.

Examples of statistical sampling include the following:

Random sampling Samples are selected at random.

Cell sampling Random selection is performed at predefined intervals.

Fixed interval sampling The sample existing at every n + interval increment is selected for testing.

No statistical Sampling

No statistical sampling is based on the auditor’s judgment (also referred to as judgmental sampling). The auditor determines the sample size, the method of generating the sample, and the number of items to be analyzed. The results of judgmental sampling are unlikely to represent the actual population. This is a subjective process usually based on elements of risk or materiality. An example of no statistical sampling includes haphazard sampling, in which the samples are randomly drawn for testing.

After the samples are selected, the next step is to perform compliance tests or substantive testing.

Conducting Audit Testing As stated earlier, the basic test methods used will be either compliance testing or substantive testing. Appropriate audit samples will have to be generated for the test.

Compliance Testing

Compliance testing tests for the presence or absence of something. Compliance testing includes verifying that policies and procedures have been put in place, and checking that user access rights, program change control procedures, and system audit logs have been activated. An example of a compliance test is comparing the list of persons with physical access to the data center against the HR list of current employees.

Compliance testing is based on one of the following types of audit samples:

Attribute sampling Generally popular in compliance testing. The objective is to determine whether an attribute is present or absent in the subject sample. The result is specified by the rate of occurrence―for example, the presence of 1 in 100 units would be 1 percent.

Stop-and-go sampling Used when few errors are expected. Stop-and-go allows the test to occur without excessive effort in sampling and provides the opportunity to stop testing at the earliest possible opportunity. It is a simple form of testing to reinforce any claim that errors are unlikely in the sample population.

Discovery sampling A 100 percent sampling used to detect fraud or when the likelihood of evidence existing is low. Forensics is an excellent example of discovery sampling. This is an attempt to discover evidence.

Precision, or expected error rate The precision rate indicates the acceptable margin of error between audit samples and the total quantity of the subject population. This is usually expressed as a percentage, such as 5 percent. To obtain a very low error rate, it is necessary to use a very large sample in testing. Auditors are justified in using a smaller sample size when the total population is expected to be error-free. A larger sample is required when errors are expected to be present in the population. The larger sample can yield a higher average. When errors are expected, the auditor must examine more data to determine whether the actual errors are within a tolerable error rate (maximum errors you would accept).

Error levels may be determined by reviewing the findings of a prior audit and by considering changes in the organization’s procedures. Use the risk-based audit strategy to determine whether your samples and tests are telling the truth about the audited.

Substantive Testing

Substantive testing seeks to verify the content and integrity of evidence. Substantive tests may include complex calculations to verify account balances, perform physical inventory counts, or execute sample transactions to verify the accuracy of supporting documentation. Substantive tests use audit samples selected by dollar value or to project (forecast or estimate) a total for groups with related characteristics.

Substantive testing is based on one of the following types of audit samples:

Variable sampling Used to designate dollar values or weights (effectiveness) of an entire subject population by prorating from a smaller sample. Consider the challenge of counting large volumes of currency by its weight. Variable sampling could be used to count currency by multiplying the physical weight of one unit by the total weight of the combined sample, and then multiplying by the face value printed on the bill or coin. A demonstration is a single $50 bill weighing 1.0 gram, with the entire sample of $50 bills weighing 61 grams altogether. The combined sample weight would indicate a total quantity of 61 bills for an estimated dollar value of $3,050. This is a common technique for forecasting quantity and value of inventory based on particular characteristics.

Unsatisfied mean estimation Used in an attempt to project an estimated total for the whole subject population.

Stratified mean estimation Used to calculate an average by group, similar to demographics, whereby the entire population is divided (stratified) into smaller groups based on similar characteristics.

Examples are teenagers from the ages of 13 to 19, people from the ages of 20 to 29, people from the ages of 30 to 39, and those who are male or female, smokers or nonsmokers, and so on. Difference estimation Used to determine the difference between audited and unaudited claims of value.

The following answers are incorrect:

The other options like Random Sampling, Cell Sampling and Fixed Interval Sampling are examples of Statistical sampling.

The following reference(s) were/was used to create this question:

CISA review manual 2014 page number 55 to 56

CISA certified information system auditor study guide Second Edition Page Number 98 to 101