AUDITORS' INFORMATION SEARCH AND EVALUATION IN HYPOTHESIS-TESTING: POSITIVITY, EXTREMITY, PSEUDODIAGNOSTICITY AND CONFIRMATION EFFECTS

This study examined in an analytical procedures context, how auditors search for information and how they evaluate the acquired information in the course of hypothesis-testing. In the information search phase, auditors' preferences in their choice of questions were examined to determine if they exhibited positivity and/or extremity. In the subsequent information evaluation phase, after receiving answers to the questions, auditors' probability judgments of the validity of their inherited (or null) hypothesis were analyzed. This was to determine if they were pseudodiagnostic in the evaluation of their inherited hypothesis, that is, whether they ignored the probability of the datum given their alternative hypothesis in arriving at a probability judgment of the validity of their inherited hypothesis. It was postulated that the combination of extremity in information search and pseudodiagnosticity in information evaluation would lead to confirmation effects in hypothesis-testing.; Auditors from the Big Six international accounting firms were given a short case of a company where one of two error hypotheses was responsible for fluctuations in the financial statement data. They were then given four questions and asked to indicate their preferences among the four questions in order to test for one of the two error hypotheses, which became their inherited hypothesis. Subsequently, they were given either a "yes" or a "no" answer to each of those questions and asked to indicate their judgments of the probability of the validity of their inherited hypothesis given each answer.; The results of this study indicate that in the information search phase, auditors preferred the positive and extreme question. In the information evaluation phase, as predicted by the pseudodiagnosticity hypothesis, auditors overestimated the probability of the validity of their inherited hypothesis when they received a "yes" answer to the positive and extreme question. This "yes" answer to the positive and extreme question supported their inherited hypothesis. The cumulative effect of the results from both phases of the study indicate that auditors have a tendency to confirm their inherited hypothesis. This result is contrary to the conclusions of previous auditing studies.