Statistical Sampling Techniques and Their Application to Inventories
BY HOMER E. SAYAD Partner, Saint Louis Office
Presented before the Saint Louis Chapter of The Institute of Internal Auditors — December 1958
Agreat deal has been written during the past few years on the application of statistical sampling techniques in the accounting
and auditing field. The growth of interest in this subject is evidenced by the fact that both the American Institute of Certified Public Accountants and The Institute of Internal Auditors have recently organized committees on statistical sampling. Industry also is giving serious consideration to the use of statistical sampling techniques in the area of accounting and auditing and some companies
are already using them.
Before proceeding further, it may be well to define the term "sampling." Sampling is the examination of a part of anything presented for inspection purporting to represent the aggregate. It may be said that this is not a new concept in the auditing field since in auditing we rely on test-checking of transactions, vouchers, payrolls,
etc., instead of total examination. Test-checking is a recognized
and accepted auditing procedure. References to it first appeared
in accounting literature in the 1890s and early 1900s. In the booklet entitled "Generally Accepted Auditing Standards" issued by the American Institute of Certified Public Accountants, frequent references are made to the application of audit procedures by tests and the "testing technique."
Recent discussion of this subject has centered not around sampling
or testing as such but rather around the methods and techniques
for selecting a sample. It is charged that accountants' selection of a sample is empirical and intuitive and not scientific. It is argued that since decisions based on sampling are actually decisions based on some kind of probability, the selection of the sample should be through the more objective and scientific methods of the statistician in which probability is a key factor.
I do not intend to discuss in great length the divergent views of accountants and statisticians as to the applicability of statistical sampling to accounting and auditing. I would like, however, to examine briefly the main area of disagreement, or perhaps I should say the main area of miscomprehension.