Audit Sampling Techniques: Statistical vs. Nonstatistical Methods Compared

The Role of Sampling in Auditing

Audit sampling is the application of audit procedures to less than 100% of items within a population for the purpose of drawing conclusions about the entire population. Sampling is necessary because testing every transaction is often impractical or cost-prohibitive. The auditor must design and select a sample that is representative of the population and apply appropriate procedures to the sampled items to form conclusions about the population as a whole.

CPA AUD candidates must understand both statistical and nonstatistical sampling approaches, the circumstances under which each is appropriate, and how to evaluate sampling results. The exam tests conceptual understanding, method selection, and the evaluation of sampling risk.

Statistical vs. Nonstatistical Sampling

Statistical sampling uses mathematical probability theory to select sample items and evaluate results. It provides a quantitative measure of sampling risk, which is the risk that the sample results lead to an incorrect conclusion about the population. Nonstatistical sampling, also called judgmental sampling, relies on the auditor professional judgment to select items and evaluate results. Both approaches are acceptable under auditing standards, provided the sample is designed to achieve the audit objective.

The primary advantage of statistical sampling is that it provides an objective, quantifiable measure of sampling risk. This allows the auditor to design the sample to achieve a specific level of confidence and to extrapolate the results to the population with a known margin of error. The primary advantage of nonstatistical sampling is flexibility; the auditor can use judgment to focus on higher-risk items or items of particular interest.

Attribute Sampling

Attribute sampling is used in tests of controls to estimate the rate of deviation from a prescribed control procedure in a population. The auditor defines the control to be tested, the deviation condition, and the tolerable rate of deviation. The sample size is determined based on the expected population deviation rate, the tolerable deviation rate, and the desired confidence level (which determines the allowable risk of overreliance).

After testing the sample, the auditor calculates the sample deviation rate and compares it to the tolerable deviation rate. If the sample deviation rate, adjusted for sampling risk, does not exceed the tolerable deviation rate, the auditor can conclude that the control is operating effectively. If the sample deviation rate exceeds the tolerable deviation rate, the auditor concludes that the control is not operating effectively and must modify the audit plan accordingly.

Variables Sampling

Variables sampling is used in substantive testing to estimate the monetary amount of a population or the monetary misstatement in a population. Common variables sampling methods include mean-per-unit estimation, which estimates the total value of the population by multiplying the sample mean by the number of items in the population; ratio estimation, which uses the ratio of audited values to book values in the sample to estimate the total audited value of the population; and difference estimation, which estimates the total misstatement in the population based on the average difference between audited and book values in the sample.

Monetary Unit Sampling

Monetary unit sampling (MUS), also called probability-proportional-to-size (PPS) sampling, is a statistical sampling method that combines attributes of both attribute and variables sampling. In MUS, the sampling unit is the individual dollar rather than the individual transaction. This means that larger dollar items have a proportionally greater chance of being selected, which is advantageous because larger items contribute more to the risk of material misstatement.

MUS is particularly effective when the auditor expects few or no misstatements in the population. Its advantages include automatic stratification by dollar amount, no need to estimate the population standard deviation, and a conservative approach that is less likely to accept a materially misstated population. However, MUS has limitations: it does not work well for testing completeness assertions (because unrecorded items have zero or no dollar value in the population), and it can produce overly conservative results when the population contains many small misstatements.

Sample Size Considerations

For both statistical and nonstatistical sampling, the auditor must determine an appropriate sample size. Factors that affect sample size include the tolerable misstatement or tolerable deviation rate (lower tolerance requires a larger sample), the expected misstatement or expected deviation rate (higher expected rates require larger samples), the desired confidence level (higher confidence requires larger samples), and the population size (though for large populations, the sample size is relatively insensitive to population size).

Nonstatistical Sampling Methods

When using nonstatistical sampling, the auditor selects items using judgment, which may include haphazard selection (selection without any conscious bias, but not random), block selection (selecting contiguous items from a specific time period or sequence), and targeted selection of specific items based on risk characteristics. The auditor evaluates the results qualitatively, considering the nature and cause of identified misstatements and whether they indicate systematic errors that could affect the population.

Evaluating Sampling Results

Regardless of the sampling method used, the auditor must project the results to the population and consider sampling risk. For statistical sampling, this projection involves mathematical extrapolation. For nonstatistical sampling, the auditor uses professional judgment to assess whether the sample results, considered together with other audit evidence, provide sufficient appropriate evidence to conclude on the assertion being tested.

CPA Exam Preparation Tips

• Know the difference between attribute sampling (tests of controls) and variables sampling (substantive tests).
• Understand the mechanics and advantages of monetary unit sampling, including when it is most and least appropriate.
• Be able to identify the factors that affect sample size and the direction of their effect.
• Understand sampling risk, including the risk of overreliance and the risk of incorrect rejection.
• Practice evaluating sampling results and drawing appropriate conclusions about the population.

Sampling is a practical skill that every auditor uses. Mastering the conceptual foundations and the mechanics of different sampling methods will serve you well on the CPA exam and in audit practice.