As per AAS 15, “Audit Sampling”, the auditor should select sample items in such a way that the sample can be expected to be representative of the population. This requires that all items in the population have an opportunity of being selected.
There are two major methods in which the size of the sample and the selection of individual items of the sample are determined. These methods are :
(1) judgmental sampling; and
(2) statistical sampling.
Whatever may be the method, judgmental or statistical sampling, the sample must be representative. This means that it must be closely similar to the whole population although not necessarily exactly the same. The sample must be large enough to provide statistically meaningful results.
Judgmental Sampling : Under this method, the sample size and its composition are determined on the basis of the personal experience and knowledge of the auditor. This method has been in common application for many years because of its simplicity in operation. Traditionally, the auditor on the basis of his personal experience, will determine the size of the sample and express it in terms that number of pages or personal accounts in the purchases or sales ledger to be checked. For example, March, June
and September may be selected in year one and different months would be selected in the next year. An attempt would be made to avoid establishing a pattern of selection year after year to maintain an element of surprise as to what the auditor is going to check. It is a common practice to check large number of items towards the close of the year so that the adequacy of cut-off procedures can also be
determined. The judgmental sampling is criticised on the grounds that it is neither objective nor scientific. The expected degree of objective cannot be assured in judgmental sampling because the risk of personal bias in selection of sample items cannot be eliminated. The closeness of the qualities projected by the sample results with that of the whole population cannot be measured because the sample has not been selected in accordance with the mathematically based statistical techniques. However, it may be stated that the auditor with his experience and knowledge of the client’s business can evaluate accurately enough the sample findings to make audit decision and the mathematical proof of accuracy in some cases may be a luxury which the auditor cannot afford.
In judgmental sampling the auditor’s opinion determines the sample size but it cannot be measured how far the sample size would fulfill the audit objective. In statistical sampling, the sample results are measurable as to the adequacy and reliability of the audit objectives.
Statistical Sampling : Statistical sampling is a method of audit testing which is more scientific than testing based entirely on the auditor’s own judgment because it involves use of mathematical laws of probability in determining the appropriate sample size in varying circumstances. Statistical sampling has reasonably wide application where a population to be tested consists of a large number of similar items and more in the case of transactions involving compliance testing, debtors’ confirmation, payroll checking, vouching of invoices and petty cash vouchers. Students may note that it is unnecessary for the auditor to gain indepth knowledge of statistics before making use of statistical sampling for audit testing since published statistical tables are available which
indicate the sample size based on pre-determined criteria.