No abstract

This paper exatiimes the probability judgments made by auditors for their tinancial statement lootnote disclosure decisions and their audit report additional paragraph decisions m the presence of material loss contingencies. In the United States these judgments are governed by SFAS No, S and SAS No, 58, Two prior studies ha\e reponed niconsistenl results pertaining lo the degree of compliance of auditors with the judgment and decision-making process implied by SR^S No ,'. In contrasi, S.^S No. 58 has not previously been e.xaniined with respect to auditor compliance. Results from an experimettl uith 64 audit partners from six tirms were consistent with the hypoihesis that audit partners employ the same (compensatory) judgment process when making footnote disclosure decisions as when tnaking additional paragraph decisions. This process led the partners to make judgments for the additional paragraph decision that were in accordance with S,'\S No, .'is. In contrast, their judgments for the lootnote disclosure decision were consistent with a decision theory tnodel of auditor behavior rather than the noncompensatory process intplied hy SFAS No, S, ,'\dditional aspects of the study penain to the additional paragraph decision oi auditors. Observed judgments tor these decisions are consistent uith the hypothesis that additional information is conveyed bevond that provided by only a lootnote disclosure. This is contrary to some recenl discussions of the standards.* The helpful comments of Ramji Balakrishnan; ,lere Francis; Bill Scott, the editor, tvio anonymous reviewers; and participants of the .Accounting Workshop at the tini\ersit> *:\i Iowa are gratefully acknowledged.Contemporary .Accounting Research Vol. 7 No, 2 pp. 544-564 SAS n" 58. Par opposition, leurs jugements relatifs a la presentation sous toime de notes infrapaginales etaient conformes a un modele de comportement du verificateur inspire dc la iheorie de la decision plutol qu'au processus non compensatoire que suppose le SFAS n" .5. D'autres aspects de l'etude portent sur la decision des verificafeurs relative a i'ajoui de paragraphed supplementaires. Les jugements observes re)ativemenl a cette decision sont conlormes ii i'hypothese selon laquelle d'autres renseignetnents som transmis, outre ceux qui sont fournis par voie de simples noies infrapaginaies. Ces re.sultats viennent contredire cerfaities analyses recentes portam sur les normes.

This paper reports on experiments which test whether factors such as regret or disappointment influence taxpayer compliance decisions. Previous tests of regret and disappointment theory have been based upon the common-ratio effect in which probabilities vary while outcomes are held fixed. In contrast, our experiments involved trials in which the outcomes were allowed to vary. Previous tests had mainly found evidence of regret effects and to a lesser extent were supportive of disappointment. In contrast, we were able to reject simple theories of both regret and disappointment. A second experiment produced evidence which indicates that whether or not rejected risky alternatives are resolved has no significant influence on subjects' choices. One would expect that regret could only occur when a subject learns the outcome of a non-chosen option. Hence this result can be seen as evidence against regret theory. We shall argue that there is a possible interpretation of regret theory which is compatible with this result.

No abstract

Previous experimental judgment research in accounting has been interpreted as supportive of the linear model as an appropriate representation of decision-making behavior in nearly all of the tasks investigated. Moreover, nonlinear models have come to be viewed as adding relatively little predictive power over that provided by the linear model, even in tasks considered to be inherently nonlinear. These conclusions were based largely upon evaluating the predictive ability of the linear model in terms of statistics measuring the proportion of variance accounted for by the model. In the present paper it is argued that, since these statistics are not independent of the experimental design, it is not clear whether the high correlations are indicative of the model's success in representing decision-making processes or instead are more the result of various features of the experimental design. It is suggested that correlational tests be supplemented with qualitative tests of the predictive ability of a model. Implications for accounting are discussed.