Discussant's response to "Using and evaluating audit decision aids";

              Discussant's Response to "Using and Evaluating Audit Decision Aids"
Stephen J. Aldersley
Clarkson Gordon
It is a pleasure to have the opportunity to comment on a paper dealing with a subject which has occupied a considerable portion of my time during the past few years—Audit Decision Aids. It should, therefore, not come as too much of a surprise that I agree with much of what is in the Ashton-Willingham paper. Throughout the paper I found myself nodding in agreement with the points being made. However, there are still a number of areas where I think the authors' efforts at organizing and categorizing the issues have led to unwar­ranted
oversimplification. I will direct my commentary to these areas. I intend to follow the basic outline of the Ashton-Willingham paper and will conclude my comments with some observations on what I perceive to be a couple of particularly difficult audit areas that just might lead to important decision aids.
The theme of this paper can probably be stated along the following lines: Decision aids have a role, ... but they need a cost benefit justification. The theme for my comments is related to the definition of decision aids adopted in this paper, i.e., "any explicit procedure for the generation, evaluation and selection of alternatives (courses of action) that is designed for practical application and multiple use.'' When you read the rest of the paper you wonder whether the authors have used a complete decision aid definition. They appear to have set up several straw men that are subsequently criticized and discredited. Would it not be appropriate to include some evaluative criteria in the definition of the decision aid? My contention is that if you don't, you may have a decision anti-aid instead. The courses of action should be "towards some well-defined objective" in a more complete definition of a decision aid.
Development Issues
One oversimplication in the paper is the distinction made between research-based as opposed to experience-based development approaches. The paper implies a dichotomy whereas, in practice, things are not nearly as simple. Although a decision aid may use a research base during development, it will not evolve into a tool solely from that perspective. The reason decision aids are often even considered is usually experience-based. Although one might argue that this is empirical research, the empiricism tends to be anecdotal rather than based on any research design. There are many examples from the past 20 or 30 years. For instance, analytical auditing and the related flow-charting technique grew out of our audit practice needs for concise system descriptions. Statistical sampling was implemented because of actual deficiencies encoun­tered
in the use of non-statistical techniques. Regression-based analytical review was introduced because of dissatisfaction with the quality of judgmental
26