Data Analytics First, Controls Second, Auditing Third: Ranking the Relative Strength of Three Fraud Detection Mechanisms

David Sunday Araoti

Abstract

Purpose: In the era of digital finance, organizations deploy multiple mechanisms to combat fraud, including Data-Driven Forensic Analytics (DDFA), Internal Control Integrity (ICI), and Continuous Auditing Systems (CAS). However, practitioners often struggle with resource allocation, unsure which mechanism yields the highest return on investment. This study addresses this gap by empirically ranking the relative strength of these three predictors in explaining Fraud Detection Effectiveness (FDE). Design/methodology/approach: The study synthesizes empirical evidence from three distinct quantitative studies conducted in Nigeria, an emerging economy with high fraud risk. Data were collected from finance and audit professionals across banking, manufacturing, public sector, and professional services (Total N=390 valid responses across studies). Multiple regression analyses were employed to determine the standardized beta coefficients ($\beta$) of each predictor. Rigorous diagnostic tests, including Variance Inflation Factors (VIF) and Harman’s single-factor test, ensured model validity and minimized common method bias. Findings: The results reveal a consistent hierarchy in predictive strength across all three datasets. DDFA emerges as the strongest predictor of FDE ($\beta \approx 0.41–0.42$), followed by ICI ($\beta \approx 0.35–0.36$), and then CAS ($\beta \approx 0.29–0.30$). The integrated models explain between 64% and 65% of the variance in FDE. This ranking suggests that while all three mechanisms are significant, analytical capability provides the greatest marginal gain in detection effectiveness, followed by structural governance, and then temporal monitoring. *Practical implications: Organizations should prioritize investments in data analytics infrastructure and skills as the primary driver of fraud detection. However, this "Analytics First" approach must be supported by robust internal controls ("Controls Second") to provide structural integrity, and continuous auditing ("Auditing Third") to ensure timeliness. A fragmented approach that neglects any of these layers will result in suboptimal detection outcomes. Originality/value: This study moves beyond establishing mere associations to providing a clear, empirically grounded ranking of fraud detection mechanisms. By synthesizing data from multiple studies, it offers robust evidence for strategic decision-making in fraud risk management, particularly in emerging economies where resource constraints necessitate prioritized investment.

Abstract

Keywords