Statistical Sampling and Its Use in the Small Engagement
by THERON L. PARR Principal, Atlanta Office
Presented before the Atlanta and Augusta Chapters of the Georgia Society of Certified Public Accountants—October 1963
WHEN accountants first began to hear of statistical sampling as a means of selecting items to be examined in audit tests I think that most of us immediately built up a mental block against the use of this technique. If you had the same reservations that I had they probably were (1) we are not mathematicians, (2) the tables, books, etc., needed to design a sample would require an extra work bag on each engagement, (3) the time required for design and selection of the sample would make it impractical, and (4) most important, it would limit or infringe upon the important factor of the auditor's judgment.
In the Fall of 1961 I was asked to participate in a training course on statistical sampling and at the end of the course to make a number of test applications on audit engagements. This session proved to me that the objections concerning the necessity for a knowledge of higher mathematics, the volume of material needed to design a sample,
and the elimination of the auditor's judgment were not valid. Subsequent use of statistical sampling techniques during the past two years has also eliminated the objection concerning time required for design and selection of the sample.
In Atlanta we are now in our third year of use of statistical sampling and to the best of my knowledge this technique is now being used to some extent on every engagement originating in this office. I think that this statement would be generally true for most offices of our Firm. We have found that the amount of time required for training our staff in the use of the plan can be measured in hours rather than days.
Why is statistical sampling considered of benefit in determining the extent of audit tests? We think the use of these techniques provides
a more objective basis for expressing the audit purposes to be accomplished by tests of transactions and account balances and a more objective basis for designing, selecting, and evaluating samples in relation to such purposes.