Shirley McKie case -
Posted: Mon Aug 07, 2006 3:19 am
Hello, I am Steve Horn the author of this weeks ‘Detail’ about the statistical foundation of the Shirley McKie case. I am a computer programmer who works in the field of statistics for industry. I know nothing about fingerprinting. I became interested in the McKie case a few weeks ago after seeing snippets of the Scottish Parliament inquiry on local TV (I live in Scotland). It seemed to me that people on both sides of the argument were of good intent. When this happens my instincts are to look for a ‘system fault’ rather than to blame individuals.
It struck me that it was very unusual for a crime, in this case perjury, to come AFTER a fingerprint match is found. Usually the crime comes first then the fingerprint work gets under way. I could also see that the statistical ‘population’ of fingerprint matches which could lead to a Shirley McKie type case is huge. All it takes is for ANY elimination identification to be disputed and for this to be repeated under oath in court. I assumed at first that a misidentification, even with faultless fingerprint work, was theoretically possible, but I can see that this explanation is not accepted in the fingerprint world.
I have been developing my ‘observation’ over the last few weeks on a web page. It basically calls into question the decision to prosecute Shirley McKie, in the light of the unusual circumstances of the case (a disputed elimination match). You might find it easier to read on my web page. I have made a few minor changes over the weekend so it is slightly different from the text in the 'Detail' section.
http://www.stevehornsc.pwp.blueyonder.c ... hooter.htm
Thanks for your interest.
It struck me that it was very unusual for a crime, in this case perjury, to come AFTER a fingerprint match is found. Usually the crime comes first then the fingerprint work gets under way. I could also see that the statistical ‘population’ of fingerprint matches which could lead to a Shirley McKie type case is huge. All it takes is for ANY elimination identification to be disputed and for this to be repeated under oath in court. I assumed at first that a misidentification, even with faultless fingerprint work, was theoretically possible, but I can see that this explanation is not accepted in the fingerprint world.
I have been developing my ‘observation’ over the last few weeks on a web page. It basically calls into question the decision to prosecute Shirley McKie, in the light of the unusual circumstances of the case (a disputed elimination match). You might find it easier to read on my web page. I have made a few minor changes over the weekend so it is slightly different from the text in the 'Detail' section.
http://www.stevehornsc.pwp.blueyonder.c ... hooter.htm
Thanks for your interest.