Maybe the misere is because the british authorities haven't taken the database size into account when the presented the evidence. It is the view of quite a few forensic scientists that the size of the database does not matter when presenting DNA evidence probabilistically. If this is the case, then taking a database size of 660,000 into account will lower the 37,000,000:1 to 56:1. And then one understands the better that things like this happens. The interested reader can check my arguments out for this in the journal Biometrics ( http://stat.tamu.edu/Biometrics/ ), issue 55 no.3 pages 671-677 (1999). Unfortunately, the paper isn't available on the net, so you'll have pay your library a visit.
Slashdot Mirror
Maybe the misere is because the british authorities haven't taken the database size into account when the presented the evidence. It is the view of quite a few forensic scientists that the size of the database does not matter when presenting DNA evidence probabilistically. If this is the case, then taking a database size of 660,000 into account will lower the 37,000,000:1 to 56:1. And then one understands the better that things like this happens. The interested reader can check my arguments out for this in the journal Biometrics ( http://stat.tamu.edu/Biometrics/ ), issue 55 no.3 pages 671-677 (1999). Unfortunately, the paper isn't available on the net, so you'll have pay your library a visit.