Audit Method: Sampling

Audit sampling as defined by ISA 530 is “The application of audit procedures to less than 100% of items within a population of audit relevance such that all sampling units have a chance of selection in order to provide the auditor with a reasonable basis on which to draw conclusions about the entire population.”

The means available to the auditor for selecting items for testing are;

• Selecting all items (100% examination)
• Selecting specific items; and
• Audit Sampling

Audit sampling can be applied using either non-statistical or statistical sampling approaches.

Statistical approach has the following characteristics:

• Random selection of the sample items; and
• The use of probability theory to evaluate sample results, including measurement of sampling risk.

 A sampling approach which does not possess the above characteristics is called non-statistical sampling.

Steps in planning the sample are as follows:

• Determining the objectives of the test;
• Defining what errors or deviation are being sought;
• Identifying the population and sampling units; and
• Deciding the size of the sample.

Generally, larger the size of the sample, the more it will be representative of a population. Many audit firms would like to standardize the size of sample based on general and specific risk factors. Errors increase the imprecision of results from sampling. Therefore, if they are expected, a larger sample size is required. Finally in evaluating misstatements, the auditor, should exclude ‘anomalous’ error(s) which are misstatement or deviation that are demonstrably not representative of misstatements or deviations in a population from his projected misstatements. However, such errors may be considered when evaluating all misstatements within the sample and auditors shall obtain sufficient appropriate audit evidence to corroborate the fact that an error or deviation is anomalous.

Practice

Auditors need to consider the specific objectives to be achieved and the combination of audit procedures that is most likely to achieve those objectives. Audit sampling is applicable to both tests of control and substantive procedures.

Audit Sample: Simulating Scientific Approach

The sampling approach in audit is practical embodiment of “reasonable assurance” concept. Steve Collings, the assessor of F8 ACCA paper, has written article, where briefly explained the main principles of sampling in audit context. In this post I would like to discuss some issues regarding sampling phenomena.

A Little Bit of Theory

Based on the distinctive characteristics of statistical sampling (random selection and use of probability theory) following classification of sampling methods can be inferred from the ISA 530:

1.      Statistical sampling:

1.1.   Random sampling;

1.2.   Systematic sampling;

1.3.   Monetary unit sampling.

2.      Non-statistical sampling:

2.1.   Haphazard sampling;

2.2.   Block selection.

Actually, there are two steps in sampling process: determining the number of items you need to select from general population, and, aftermath select those items appropriately to make sample representative. By “representative” I mean that sampled population has to possess all qualities pertained to general population. 

History

Power in his book “Audit Society: Rituals of Verification” provides some insights on the sampling application in early years of audit. Block-tests – the form of proto-sampling – was used in late XIX century to test relevant transactions within particular month. However, the scholar notes that such method “lacked statistically precise notions of representativeness”.

Normal probability distribution 

Power states that American audit was pioneer in application statistically reasoned sampling methods. The process of formalisation and development of statistical methodology in audit began in 1930s-1940s. The drivers of these methods application were pragmatism and economic focus of US firms.

So at the moment we (auditors) have to be grateful to mathematical statistics and more specifically to probability theory. Obviously, now unlike before our pursuit to be more efficient and profitable backed up scientific or semi-scientific methods and strengthened by audit standards and legislation.

Practical Perspectives

However, working in audit firm you do not have to be a mathematician to apply variety of samples. Nowadays in most cases audit firms use sampling software. To be successful in running sample process following items need to be entered into program:

·      Risk level, which is usually significant, moderate or low;

·      Number of items in general population;

·      Total value of general population;

·      Total value and number of key material items, which need to pre-selected for separate testing in addition to test of sampled items.

Risk level is essential for determining right number of items in sample, the mistake could result overaudit or wrong audit opinion.

We should not forget about the judgment in audit profession. Of course, the process of calculation the number of sampled items is rather formalised, but judgment could be used in setting up initial risk per flow of transactions, establishing key material items, utilising selection process (e.g. systematic vs. haphazard). If the judgment is used it would be necessary to document appropriately the reason and factors prompting it. This is especially relevant in cases when the judgmental approach substantially differs from what is reasonably expected in similar circumstances.