The risk matrix: uncertain results?

Abstract The consequence/likelihood risk matrix is a widely used risk analysis and reporting technique, often incorporated into risk and safety management software, and often relied on to help make decisions about the tolerability of risk. However, decision makers may not be aware of matrix design features that give rise to uncertainty, or of errors in its use causing further uncertainty. This paper builds on prior evidence to argue that poorly designed or inappropriately used risk matrices result in an increase in uncertainty and, potentially, adverse effects on people and organizational objectives. It is further argued that risk matrices should be designed for reporting as part of an overall risk assessment, and not used on their own. The evidence is reframed as a goal tree to show where design and use of the matrix might reduce, but not eliminate, uncertainty in risk assessments. Suggestions are given for reducing uncertainty in the design and use of risk matrices that may enable more effective risk assessments. The article concludes that (1) risk assessors need better training in risk matrix design and use, and (2) decision makers need training to better understand the inherent unreliability of any risk matrix as a decision support technique.