Internal control effectiveness : a clustering approach

Purpose: Internal control frameworks provide only broad guidance concerning internal control concepts, leaving the details to the adopting firms. In this study the adopted internal control structure and effectiveness in firms is examined and visualized, and a typology of firms is presented. Control structure and effectiveness are measured based on the assessment of management, rather than using reported material weaknesses as most studies do. This type of evaluation is more purposeful for firms that do not apply the Sarbanes-Oxley Act. Design/methodology/approach: The survey data (from 741 CEOs) are clustered using the self-organizing map (SOM), a visual artificial neural network approach. A three-dimensional effectiveness proxy is used. Findings: The analysis reveals four alternative types of internal control effectiveness in firms and visually presents how the components of the internal control structure (COSO) are associated with each one. A typology of internal control structure and effectiveness is then created. Practical implications: The findings suggest that there are interrelated, but not straightforward, relationships between internal control variables, and that there is a link between some of them and higher internal control effectiveness in practice. These findings have important implications for those responsible for improving or assessing internal control, such as management, personnel, and internal and external auditors. Originality/value: This article uses a clustering approach to create a typology for alternative types of internal control structure and effectiveness, based on data from actual firms. Instead of using material weaknesses as a measure, this study uses managers’ own assessments of internal control effectiveness