A clustering and visualisation approach to the analysis of internal control structures

The purpose of this study is to use a visual clustering approach (the self-organising map) to analyse differences in companies' internal control (IC) structures, based upon the five components of the Committee of Sponsoring Organizations of the Treadway Commission (COSO) framework. The empirical data were collected through a survey of 741 Finnish companies. Over 90% of the respondents were chief executive officers (CEOs) and the response rate was 50.4%. Five different clusters with differing IC structures were identified. Three of these clusters are partly in contradiction with the literature (high risk assessment vs. undeveloped control environment; advanced control environment vs. undeveloped other control structures; advanced control structures vs. low risk assessment). This study, therefore, raises several questions for future research and contributes to earlier IC and self-organising map (SOM) studies.