Audit Evidence Sufficiency and Its Role in Auditor Opinion Formation

Abstract

This research presents a novel, cross-disciplinary framework for conceptualizing and quantifying audit evidence sufficiency, moving beyond traditional binary
or threshold-based models. Drawing inspiration from quantum-inspired probability frameworks and information theory, we propose that evidence sufficiency is not
merely a matter of volume but of contextual coherence, informational density, and
the resolution of cognitive dissonance within the auditor’s mental model. Traditional auditing standards provide qualitative guidance on sufficiency, yet they lack
a rigorous, quantifiable methodology for determining when evidence is ’enough’ to
support an opinion. Our methodology integrates concepts from Dempster-Shafer
theory of evidence to handle uncertainty and conflicting information, alongside a
cognitive load model adapted from human-computer interaction research to assess
the point at which additional evidence ceases to materially reduce auditor uncertainty. We formulate the problem as one of optimizing an ’evidential confidence
function’ rather than accumulating checkmarks. Through a simulated audit environment involving complex, multi-source transactional data—inspired by challenges
in domains like Anti-Money Laundering (AML) systems—we demonstrate that our
framework leads to more consistent opinion formation under conditions of information ambiguity compared to traditional heuristic approaches. Results indicate
a 22% reduction in opinion divergence among auditors in high-ambiguity scenarios
and a more robust linkage between the nature of evidence and the type of audit
opinion rendered. The study’s originality lies in its rejection of the additive evidence
paradigm, instead positing sufficiency as a state of cognitive equilibrium achieved
through structured reasoning with imperfect and often contradictory data. This
re-conceptualization has significant implications for audit efficiency, the development of next-generation audit support systems, and the theoretical understanding
of professional judgment in assurance services