Question

Non-statistical sampling should not be used in auditing as it is not as reliable as statistical sampling. Do you (dis)agree with the statement and why?

Answer 1

There are two types of sampling - statistical sampling and non-statistical sampling. The sampling risk of a statistical plan can be controlled and measured, and on contrary even a perfectly designed non-statistical sampling would not provide for the measurement of sampling risk. I agree with the statement that non-statistical sampling is not as reliable compared to statistical sampling and should not be used in auditing. The non-statistical sampling considers all the aspects of risk in audit that are not due to sampling. An auditor may use a procedure to all transactions or balances and still may not be able to detect a material misstatement. Moreover it includes the possibility of selecting the procedures of audit that are not appropriate to achieve the specific target. For example, confirming recorded receivables may not be relied on to inform on the unrecorded receivables. Also the auditor may not be able to recognize misstatements considered in documents that he examines, which would make an ineffective procedure even if auditors were to examine all items. The statistical sampling methods would provide a quantified estimate of sampling risk, which non statistical methods do not