Auditors' Professional Skepticism: Neutrality versus Presumptive Doubt

T.L.C.M. Groot, L.M. Quadackers, A. Wright

Research output: Contribution to JournalArticleAcademicpeer-review


Although skepticism is widely viewed as essential to audit quality, there is a debate about what form is optimal. The two prevailing perspectives that have surfaced are "neutrality" and "presumptive doubt." With neutrality, auditors neither believe nor disbelieve client management. With presumptive doubt, auditors assume some level of dishonesty by management, unless evidence indicates otherwise. The purpose of this study is to examine which of these perspectives is most descriptive of auditors' skeptical judgments and decisions, in higher and lower control environment risk settings. This issue is important, since there is a lack of empirical evidence as to which perspective is optimal in addressing client risks. An experimental study is conducted involving a sample of 96 auditors from one of the Big 4 auditing firms in the Netherlands, with experience ranging from senior to partner. One of the skepticism measures is reflective of neutrality, the Hurtt Professional Skepticism Scale (HPSS), whereas the other reflects presumptive doubt, the inverse of the Rotter Interpersonal Trust Scale (RIT). The findings suggest that the presumptive doubt perspective of professional skepticism is more predictive of auditor skeptical judgments and decisions than neutrality, particularly in higher-risk settings. Since auditing standards prescribe greater skepticism in higher-risk settings, the findings support the appropriateness of a presumptive doubt perspective and have important implications for auditor recruitment and training, guidance in audit tools, and future research.
Original languageEnglish
Pages (from-to)639-657
JournalContemporary Accounting Research
Issue number3
Publication statusPublished - 2014


Dive into the research topics of 'Auditors' Professional Skepticism: Neutrality versus Presumptive Doubt'. Together they form a unique fingerprint.

Cite this