An investigation of the impact of source reliability, positivity, and order effects of evidential matter on auditors' decision processes

This study empirically tests three factors and their interactions that may affect auditors' decision processes: (1) source reliability of information, (2) positivity of source information (whether information supports or disaffirms an audit assertion), and (3) order of information presentation. It is posited that these factors impact both subordinate auditors' decisions and their information search processes. The impact on the auditors is predicted to be a multiplicative function of the three factors based, in part, on the application of social impact theory (Latane 1981). The hypothesized impact on the auditors' decisions is measured by the subjects' determination of the appropriateness of the account balance under investigation. The propensity of auditors to follow a confirmatory search strategy measures the impact on the search process.; A computer-interactive task was developed to trace the search processes of the audit participants and record their decisions. Subjects first evaluated background information about a hypothetical client and developed initial hypotheses about the adequacy of company's allowance account. The participants then searched for and evaluated evidential cues. This was followed by a final decision about the adequacy of the account.; The results of the experiment provide mixed support for social impact theory. Strong evidence was provided which indicates that negative source information greatly influenced auditors' decisions and their information search processes. In addition, the order in which source information was presented to auditors had a significant impact on the auditors' initial hypotheses although it did not induce auditors to follow confirmatory search strategies.; The reliability of source information had little impact on the auditors' decisions and their information search processes. Although the finding is contrary to the hypothesized effect of source reliability, it does support two prior studies (Joyce and Biddle 1983; Margheim 1986) which suggest that auditors are insensitive to the objectivity of source information.; The posited multiplicative effects of the three factors were generally not supported. Contrary to the predictions of the interactive hypotheses, the impacts on auditors' decisions and their decision processes, specifically the propensity of auditors to engage in confirmatory information search strategies, were not statistically significant.