Discussant's response to a decision theory view of auditing;

              Discussant's Response to
A Decision Theory View of Auditing
James K. Loebbecke
Touche Ross & Co.
Compliments are in order for Bill Felix on a fine paper. It covers the subject well and reflects elements of both tact and wisdom. Tact is evident in that it presents a model which expresses decision criteria in terms of a payoff matrix instead of a loss function. This is a distinction which, I can assure you, is particularly appealing to practicing independent public accountants. Wisdom is reflected in that it discusses several of the broader aspects of using a decision model in auditing as well as the technical characteristics of the model itself.
Within the past two years I have become increasingly involved in the challenging problem of "modeling the audit." It would seem so nice to have the complete audit model—the ultimate audit tool. My research has disclosed models which are variously described as probabilistic, stochastic, analytical and simulation models.*
My intent today is not to debate the technical details of Bill Felix' model or any of these others. Rather, I would like to consider some questions about audit models in general: are they feasible, are they desirable, and how should they be implemented?
Feasibility
Audit models are clearly feasible. This statement lies partially in the defini­tion
of audit models. Consistent with Bill's paper (which presents a model, not the model) the classical statistical inference model now common in auditing is an audit model. Other models are more complex, but none of their authors show an absence of conviction about their ultimate feasibility. Further indication of feasibility is suggested by successful applications of models in other fields. Examples of such models can be found in engineering, medicine and other sciences. Problems of computation and volumes of data previously deemed overwhelming have been successfully solved with computer assistance. This tool is causing a significant change in auditing and is the key to further advanced techniques.
* See the following examples:
William R. Kinney, Jr., " A Decision Theory Approach to the Sampling Problem in
Auditing," University of Iowa Working Paper Series No. 74-4, March, 1974.
John Neter and Seongjae Yu, " A Stochastic Model of the Internal Control System,"
University of Illinois at Urbana-Champaign Faculty Working Paper No. 106, April 1973.
Barry E. Cushing, " A Mathematical Approach to the Analysis and Design of Internal
Control Systems," The Accounting Review, January 1974, pp. 24-41.
James K. Loebbecke and David Burns, "Computer Simulation of Internal Control Systems,"
unpublished paper.
72