Which of the following statistical sampling approaches is the most appropriate for testing a population...

Discovery sampling is a statistical sampling method that is specifically designed for detecting fraud or other irregularities. It is most appropriate when the auditor expects that deviations or fraud may be rare but significant if found.

Discovery Sampling:

Discovery sampling is used when the auditor is trying to identify at least one occurrence of a particular event, such as fraud. The sample is designed so that if a single error is found, it suggests that more may exist within the population, warranting further investigation.

Application in Fraud Detection:

Discovery sampling is effective in fraud detection because it focuses on identifying whether any instances of fraud exist within a population. This approach is well-suited for situations where even a small number of fraudulent transactions could have a significant impact.

IIA Practice Guide on Statistical Sampling:

The IIA suggests that discovery sampling is appropriate when the goal is to find the presence of an error or fraud, particularly in populations where such occurrences are expected to be infrequent.

Option B (Stop-or-go sampling): This method is used to control the risk of over-auditing when errors are expected to be low, but it is not specifically designed for fraud detection.

Option C (Haphazard sampling): This is a non-statistical sampling method and is not appropriate for systematic fraud detection.

Option D (Stratified attribute sampling): This method divides the population into subgroups but is not specifically aimed at discovering fraud.

Detailed Explanation:Why Not Other Options?Conclusion: Option A is correct because discovery sampling is the most appropriate statistical method for testing a population for fraud, as it is designed to detect even a small number of significant deviations, consistent with IIA guidance.