Audit Sampling: A CPA Exam Essential
Audit sampling is a core topic on the AUD section of the CPA exam. AU-C Section 530 (AICPA) and AS 2315 (PCAOB) establish the standards for audit sampling. Sampling allows auditors to draw conclusions about an entire population by testing a subset, but it introduces sampling risk - the risk that the auditor's conclusion based on a sample differs from the conclusion reached if the entire population were tested.
Types of Sampling Risk
There are two types of sampling risk that CPA candidates must understand:
| Risk Type | Tests of Controls | Substantive Tests |
|---|---|---|
| Risk of over-reliance / incorrect acceptance | Auditor concludes controls are effective when they are not | Auditor concludes a balance is not materially misstated when it is |
| Risk of under-reliance / incorrect rejection | Auditor concludes controls are ineffective when they are effective | Auditor concludes a balance is materially misstated when it is not |
The risk of over-reliance (for controls) and incorrect acceptance (for substantive tests) affect audit effectiveness - these are the dangerous risks. The risk of under-reliance and incorrect rejection affect audit efficiency - they lead to unnecessary additional work but not to an incorrect opinion.
Statistical Sampling Methods
Statistical sampling uses mathematical probability theory to determine sample size and evaluate results. The two main categories are:
Attribute Sampling (Tests of Controls)
Attribute sampling tests the rate of deviation from a prescribed control procedure. Key elements include:
- Expected deviation rate: The anticipated rate of deviations based on prior experience or a preliminary sample.
- Tolerable deviation rate: The maximum rate of deviations the auditor will accept and still rely on the control. Typically ranges from 2% to 7%.
- Confidence level: Usually 90% or 95%. Higher confidence requires larger samples.
If the sample deviation rate exceeds the tolerable deviation rate, the auditor cannot rely on the control and must increase substantive testing.
Variables Sampling (Substantive Testing)
Variables sampling estimates the dollar amount of a population or the dollar amount of misstatement. Common approaches include:
- Mean-per-unit (MPU): Estimates the average audited value of items in the sample and projects it to the population. Requires the largest sample sizes but does not require recorded book values.
- Difference estimation: Estimates the average difference between audited and book values and projects it. Effective when differences are expected and the population is large.
- Ratio estimation: Uses the ratio of audited to book values in the sample to estimate the total audited value. Works well when misstatement amounts are proportional to book values.
Probability-Proportional-to-Size (PPS) Sampling
PPS sampling (also called monetary unit sampling or dollar unit sampling) is widely used for substantive testing of account balances. Each dollar in the population has an equal chance of selection, meaning larger items have a higher probability of being selected. Key characteristics:
- Automatically stratifies the sample toward larger items
- Individually significant items (those exceeding the sampling interval) are automatically selected
- Does not require knowledge of the standard deviation of the population
- Uses attribute sampling tables to determine sample size
- Conservative - tends to overstate the projected misstatement
- Not suitable when understatement is a concern (zero or negative balances are not sampled)
Non-Statistical Sampling
Non-statistical (judgmental) sampling uses auditor judgment rather than probability theory to select and evaluate the sample. It is equally acceptable under auditing standards, but the auditor cannot quantify sampling risk. CPA exam questions often test the distinction:
- Non-statistical sampling still requires a representative sample and proper documentation.
- The auditor must still define the population, determine an appropriate sample size, and project misstatements found in the sample to the population.
- Selection methods for non-statistical sampling include haphazard selection (not the same as random) and block selection.
Sample Selection Methods
Regardless of whether sampling is statistical or non-statistical, the auditor needs a selection method:
- Random selection: Every item has a known, equal probability of selection. Required for statistical sampling.
- Systematic selection: Select every nth item after a random start. Risk of bias if the population has a pattern.
- Haphazard selection: Auditor selects without a structured technique but attempts to be unbiased. Acceptable only for non-statistical sampling.
- Block selection: All items in a selected block (e.g., all transactions in March). Generally not appropriate as a sole selection method.
Evaluating Sample Results
After testing the sample, the auditor must project the results to the entire population. For substantive tests, this means calculating the projected misstatement and comparing it to tolerable misstatement. If projected misstatement approaches or exceeds tolerable misstatement, the auditor should consider expanding the sample, performing alternative procedures, or proposing an adjustment.
Strengthen your understanding of audit sampling by working through CPA flashcards and testing yourself with timed mini exams. Sign up free to get started.