Evaluation and reporting of scientific evidence: the impact of partial probability assignments

The assessment of the value of scientific evidence can be performed by the derivation of a likelihood ratio, a rigorous concept that provides a measure of the change produced by an item of information in the odds in favor of a proposition as opposed to another. This represents a demanding task with several sources of uncer- tainty, due for example to elicitation of probabilities or to computational impasses. While use of such a metric is well established and supported by operational stan- dards, opinions about what should be an appropriate way to deal with such sources of uncertainty while presenting expressions of evidential value at trial differ. Some quarters promote positions according to which practitioners should state a range of values for the probabilities of the evidence given competing propositions, and report a range of values for the likelihood ratio. However, such partial probability assignments may not make good use of available information.