Reply.

The authors thank Barry Lewis and Dan O' Leary for their thoughtful comments on the author's research. Lewis states the case for experimental testing of computational testing of computational models very well. The GCX model proposes that a key component of auditor expertise is client-based knowledge. Auditors have detailed knowledge of world events that affect their clients' operations and finances. Auditors use this knowledge extensively when they make going-concern judgments and the GCX research provides a model of how that process takes place. The development of valid computational models can make significant contributions to the knowledge of auditor judgment. A clear distinction needs to be made between computational models that will be used in practice as a decision aid and those that are proposed as cognitive models to be used by the research community. Scientific knowledge of cognition in auditing is, at present, limited. This state of scientific knowledge extends to human cognition in general.