An empirical examination of auditors' assertion-level inherent risk assessments: Tests of explanations for constancy over assertions.

Abstract

The audit risk model is central to audit planning and the allocation of resources during audits. The auditing standards (AICPA 1994) and the practices of large accounting firms indicate this. However, practical application of the model is very complex. Two recent papers deal with a particular aspect of this complexity as it relates to the risk of misstatements in assertions and the assessment of this risk. Kinney (1992) proposes the existence of dependence in the risk of misstatement between groups of two or more financial statement components (e.g., assertions) and Waller (1993) presents data consistent with that proposition. In addition, Waller (1993) presents data which displays a substantial degree of dependence in risk assessments over assertions within an account. The current study develops and empirically tests three potential explanations for this latter finding. Despite the importance of risk assessment to audit planning and auditing in general, limited research of an experimental nature exists (Colbert 1988; Daniel 1988; Jiambalvo and Waller 1984; Libby, Artman, and Willingham 1985; Strawser 1991). None of these studies deal with risk assessments made at the assertion level. Given a policy of assessing risk at the assertion level, increased understanding about assertion-level inherent risk assessments is important. Specifically, given dependence in assertion-level risk assessments, identifying sources of this dependence is important to a better understanding of auditors' risk assessments. The first explanation involves the extent to which pairs (or larger groups) of assertions are audited using similar procedures. The second explanation involves the extent to which auditors focus on a single assertion as most important (i.e., the use of a "Most Important Assertion Heuristic"). It was hypothesized that a relationship exists between the extent to which a single assertion is considered most important and the dependence in inherent risk assessments over assertions. The third explanation hypothesized that dependence in inherent risk over assertions is related to the extent to which auditors focus on "general" as opposed to "assertion-specific" risk factors. Neither of the first two explanations received empirical support. The third explanation received very limited empirical support.