Statistical sampling has been widely used in various fields of industry, medicine and marketing. The successful use of statistical sampling in the mentioned fields have stirred the interest of the independent auditors who normally use sampling technique. Statistical sampling improves efficiency and reliability of the audit test because judgement sampling which is normally applied by auditors lacks objectivity, does not provide a reliable basis for sampling and is dependent upon personal opinions and experiences which vary from one auditor to another. The use of statistical sampling is permitted under generally accepted auditing standards by relating measurement of sampling risk which comprises precision and confidence level to audit objective. As statistical sampling technique is not yet widely used in auditing and auditors generally are not familiar with it, a research into its application and problems which may be encountered will be useful to most auditors. In writing this thesis, the author obtained information from various text books, the opinion of prominent accountants and the American Institute of Certified Public Accountants (AICPA), from researches carried out by the scholars, from the audit manual of the firm where the author is currently working, and from the author own experience gained directly from the audit field work. As this topic has never been seriously studied before in Thailand, most of the information was therefore obtained from research papers compiled abroad and particularly from the opinion in the special report issued in 1962 by the committee on statistical sampling assigned by AICPA.
The method of study in this thesis traces through the basic sampling problems which consist of determination of test objective, selection of sample, determination of sample size and evaluation of sample result including decisions made by auditors from sample result. In the chapter before last, examples are given of application of statistical sampling to various audit tests performed by the auditors. Although statistical sampling has more advantages than judgement sampling, it does not imply that auditors have to apply statistical sampling in every audit test. In certain circumstances judgement sampling may be more appropriate. There are some problems and limitations in the application of statistical sampling to audit tests such as the problems of relating precision and confidence level to audit objective, audit cost and knowledge of statistical methods etc. However, it is useful for auditors to know the method of application of statistical sampling as this will give them the opportunity to consider which circumstances statistical sampling is more appropriate. In addition, when the business enterprise is already using sampling e.g. where the internal audit department uses statistical sampling, the external auditor should be competent to evaluate the reliability and validity of the method used by the business enterprise. In the near future, it is likely that statistical sampling will play an important role in auditing because most of the business operations will increase in size and the volume of data to be audited by the auditors will also increase. Auditors will need a more reliable and scientific method of sampling rather than judgement sampling which is currently used by most auditors.