Jane Dillard-Eggers and Michael L. Roberts
Advances in Accounting Behavioral Research, Vol 13, pp.89-111

In light of advances in the theory of cognition (Anderson, 1996, 2000; Anderson & Fincham, 1994; Anderson & Lebiere, 1998) and research on learning from worked examples (Atkinson et al., 2000; Cooper & Sweller, 1987; Sweller & Cooper, 1985), this study extends earlier research findings that auditors need practice and certain kinds of feedback to acquire procedural knowledge to identify causes of variations between expected and actual financial ratios. We test an alternative form of instruction: worked examples. As predicted by Anderson’s ACT-R 4.0 theory, the results indicate individuals’ pre-test declarative knowledge interacts significantly with learning method (with or without examples) on procedural knowledge acquisition. In contrast to prior findings, this study shows that improvements in auditing procedural knowledge can be achieved by passive instruction in worked examples, a potentially more efficient (cost-effective) method than practice and feedback for auditor training.