exploratory sampling to assure that the proportion of units with a particular attribute (i.e., error) is not in excess of a given percentage of the population. Three determinations needed to use discovery sampling are: (1) size of population; (2) minimum unacceptable error rate; and (3) confidence level. Sample size is provided by a sampling table. If none of the random samples has an error, the auditor can conclude that the actual error rate is below the minimum unacceptable error rate. Usually, discovery sampling is employed to identify batches of documents requiring detailed examination. Assume the auditor desires to determine the correctness of costing of documents from 20 branches. A discovery sample can uncover those batches having, for example, a 95% probability of an error rate below 1%. The auditor will accept those batches as satisfactory and examine in detail the remaining batches. It is a good procedure to follow in checking the quality of clerical work. When limited time exists, discovery sampling can reassure the auditor that the error rate is less then a certain percentage using a small sample size. Discovery sampling can also be used to test audit reliability ex post. Assume an error was not detected when the CPA used a random sample in examining the population. However, after the error is uncovered, the auditor can determine the probability of having found this error. The CPA may have checked random units giving him a 95% confidence level that the error rate in the population was below 1%. The incorrect units are, say .1% of the population, hence the method and assumptions used were appropriate. Assume the auditor looks at an inventory list comprised of quantity, unit cost, and total cost. It would be impractical for auditors to examine each pricing and extension. They can utilize discovery sampling to derive a 90% confidence level that the error rate in pricing and extension is below 1%. The table indicates that for 2000 inventory items, a random sample size of 220 is appropriate. If no errors are found, the accountant concludes the entire inventory list is correct. If one error is uncovered, the accountant ceases sampling and checks all of the extensions on the list. A problem with discovery sampling is the rejection of some acceptable batches. While discovery sampling may be used by the internal auditor as a final check, the external auditor should use it only as a preliminary scanning procedure to test the quality of data in a population.
in statistics, exploratory sampling to assure that the proportion of units with a particular attribute (i.e., error) is not in excess of a given percentage of the population. Three determinations needed to use discovery sampling are: (1) size of population; (2) minimum unacceptable error rate; and (3) confidence level. Sample size is provided by a sampling table. If none of the random samples has an error, the auditor can conclude that the actual error rate is below the minimum unacceptable error rate.
