Addressing the bias in probability and impact assessments

Most Project Managers will use a probability and impact grid (or matrix) in some form as a means to assess the relative importance of identified risks and hence decide which risks need urgent treatment or, in some rare cases, no treatment at all. For many, either subconscious or conscious bias will mean that they are wasting their time because of the way people tend to think about threats and opportunities. The majority of those who use a probability and impact grid use the ‘simple’ 3 x 3 grid with axis scales of high, medium and low. Unfortunately, there is very often no concept of what is meant by high, medium or low probability or what is meant by impact. Is it impact on the whole project or impact against the schedule, the budget, the scope, the quality, the expected outcomes (i.e. benefits) or what?