2.6 Communicating uncertainty

Accurately communicating uncertainty in forensic science is crucial in order for the fact finder to fully assess the evidence. There is consensus from the UK Forensic Science Regulator and ENFSI that a likelihood ratio, range of likelihood ratios, or an equivalent verbal statement should be used to communicate direct uncertainties about the evidence. Examples include:

• single number: the recovered number of indistinguishable fibres provide 50 times more support for the hypothesis that the questioned activity took place in the living room compared to the bedroom.
• range of numbers: the recovered number of indistinguishable fibres provide between 10 to 100 times more support for the hypothesis that the questioned activity took place in the living room compared to the bedroom.
• verbal qualifier: the recovered number of indistinguishable fibres provide moderate support for the hypothesis that the questioned activity took place in the living room compared to the bedroom.

The resulting statement is always relative, in that it gives the fact finder information about the scientific evidence in light of the prosecution’s assertions versus those of the defence.

Indirect uncertainties in forensic evidence are usually given as a subjective verbal statement, or a factual statement describing the strengths and weaknesses of the quality of the evidence. No systematic approach is currently used for assessing and communicating indirect uncertainties. In the medical field, the GRADE scale is routinely used to assess the effects of medical interventions. This scale has categories which range from ‘very low quality’ to ‘high quality’ depending on how the evidence compares against a standardised set of characteristics.

Assessing and communicating uncertainty plays an important role in evaluating forensic evidence. It is important to use a standardised framework for this so that any assumptions and conclusions that are made are clear and assessable for the court. The mathematical tool of probability theory provides this framework and ensures a common language for logical inference and is the focus of the next chapter.