The Audit Risk Model: An Empirical Test for Conditional Dependencies among Assessed Component Risks

Professional standards and prior theoretical research indicate that assessed audit risk components should be conditionally dependent. In an experiment, experienced auditors made the risk assessments that are, in practice, inputs for using the audit risk model for planning the extent of detailed testing. Conditional dependencies were tested using a sequential linear modeling process that added the previously assessed risk components to the model (e.g., inherent risk assessments added to predict subsequent control risk assessments) as the last independent variable. Results showed that the previously assessed risk substantially increased the explanatory power of the models in accounting for variation in the subsequently assessed components. The results support the notion that audit risk components are assessed conditionally. Thus, they provide a defense for practitioners' claims that they are appropriately using the model and give guidance to future research on the audit risk model.