Summary of uncertainty characterization methods

Methods available for analysing uncertainties fall into three broad types qualitative, deterministic and probabilistic. They provide contrasting ways of characterizing the relative importance of the uncertainties affecting an assessment and of characterizing the overall uncertainty of the assessment output, and they provide an essential input for decision-making. 

In general, probabilistic approaches require more expertise and time than deterministic and qualitative approaches. For this reason, it is efficient to adopt a tiered approach, which starts by considering all uncertainties qualitatively (Tier 1). This may be sufficient, if the outcome is clear enough for risk managers to reach a decision. Otherwise, the uncertainties that appear critical to the outcome may be analysed deterministically (Tier 2) or probabilistically (Tier 3). 

A range of alternative methods exists at each tier, each with its own advantages and disadvantages. Although there is substantial experience with uncertainty analysis in some fields (e.g. climate change), it would be premature to make prescriptive recommendations on which methods to use in exposure assessment. For example, when discussing the use of probabilistic methods for microbial risk assessment, the former European Commission Scientific Steering Committee concluded that a quick harmonisation at the present state-of- 

---