An Empirical Investigation of the Relation between Risk Assessments and Sample Size Decisions

This study examines the relation between auditor risk assessments and sample size decisions for a sample of 235 audit sampling applications from 64 audits of medium-sized companies performed by three large accounting firms. Auditors relied on controls and assessed inherent risk below the maximum on most audits. Although sample sizes are associated with both inherent risk and control risk assessments, the relationship was significant for only one of the three audit firms. We find evidence that risk assessments are lower when there is a relationship between risk assessments and sample size decisions, consistent with the concept of auditors working backwards to reduce required sample sizes. The firms in our study also differed in average sample sizes and number of risk assessment levels. This suggests that sensitivity of audit effort to risk may depend upon the way in which risk is assessed, and overall sample size levels