Decision Trees Versus Decision Tables for Audit Test Data Design: An Experimental Study*

Use of test data is an important evidence collection method that auditors can employ to evaluate the authenticity, accuracy, and completeness of program processing. Nevertheless, test data design is a time-consuming and error-prone process. To facilitate design, two approaches advocated are the decision-tree approach and the decision-table approach. An experiment conducted to examine the quality of test data designs produced using both approaches showed some support for the superiority of the decision-table approach over the decision-tree approach. Insofar as the results obtained are problematical, however, the research needs to be replicated and extended.