دانلود مقاله قضاوت حرفهای و احتیاط در استفاده از تکنیک های نمونه برداری
Abstract
References
Abstract
Audit sampling means, according to International Standard on Auditing 530 “Audit Sample”, applying audit procedures to less than 100% of the items or class of transactions within an account balance. Sampling is used not only in auditing financial statements, but widely in market research, scientific analysis, market analysis, surveys. An ideal situation would involve studying the entire population under investigation. This is impossible in situations where we find large populations of data. The auditor must use professional judgment to assess audit risk and establish appropriate procedures for the transactions and accounts tested. When the auditor uses sampling, his goal is to ensure that the sample provides a reasonable basis to draw conclusions about the population from which the sample is selected. Using statistical sampling assumes a computer program, more expensive, it requires statistical knowledge and assumes in a lesser extent the use of professional judgment. Non-statistical sampling does not allow quantification of risk, leaving large margin of interpretation, exposing the auditor to a high risk of malpractice.
Introduction
Researchers all over the world have begun to question the sampling techniques applied, underlying its disadvantages and risks.
John Wendell and Josef Schmee (Wndell and Schmee,1996, pp.825) noted that statistical sampling auditing methods can create problems that do not apply to a large sample of items. Populations are usually small and the number of errors accepted, even in the case of a large sample of elements, is almost zero, setting aside the central limit theorem. Central Limit Theorem (CLT) is a statistical theory, which argues that, given a sufficiently large sample from a population diverse enough, the average of all results will be approximately equal to the average population. In addition, evidence will follow approximately a normal distribution model (eg, bell-shaped curve), with variations that reflect the source population variance divided by sample size. As a rule, a sample of 50 elements is necessary for the CLT to be applied.