How Does an Initial Expectation Bias Influence Auditors' Application and Performance of Analytical Procedures?

Prior research demonstrates that knowledge of unaudited balances biases auditors' expectations during analytical procedures. What is less understood is how these biases affect auditors' subsequent investigations and their conclusions about the reasonableness of a particular balance. We employ the selective accessibility model to examine the differences in analytical procedure performance when auditor expectations are formed with versus without knowledge of the client's unaudited financial statement balances. In an experimental setting, we found that auditors with knowledge of unaudited balances favored hypotheses and supporting information indicating that the client's balance was reasonably stated. Auditors who formed expectations without current-year figures were more willing to evaluate competing alternatives, could better identify the most pertinent information, and were significantly more likely to identify a material misstatement using an analytical procedure.