Crown Prosecution Service seeks to limit the scope of a Court of Appeal decision on likelihood ratios
Issue 70 of this e-wire considered
the case of R -v- T in which the Court of Appeal looked at likelihood
ratios. It concluded that justice is not served by dressing up an expert’s
guesses as pseudo-science.
This case led the court to be critical of the reliance placed on an
insufficiently large database, the Forensic Science Services’ Footwear Database
(FD). Since the judgment in R -v- T, the Crown Prosecution Service (CPS)
has published some troubling guidance on this matter.
pointing out rightly that the ‘ruling only applies to evidence based on the
Footmark Database’, goes on to assert that the ‘principles should not, at this
stage, be applied to other database analysis’. This is very worrying.
Of course, the
specific criticism of the FD relates just to that database. But the criticism
has a general nature as well as a specific one. Any database that is too
small to support the kind of statistical manipulation that was used on the FD
should be caught by the principles set out in the R -v- T judgment, and
the CPS should have the wit to acknowledge this general application.
can be found by searching Google for Footwear Database Bayesian statistics.