Jeff Jonas has written a great blog post here….
on the ‘false positives’ and ‘false negatives’ that ALL databases produce, but in particular, focussing upon forensic DNA databases. He suggests that such databases should be ‘self-correcting’, so that any new data entered that contradicts prior data generated, should immediately be flagged up. So, if a DNA ‘match’ had convicted an innocent person, yet a later DNA sample entered into the database also matched the crime scene for the earlier conviction, then an alert would be created. Jonas looks at the CODIS system and makes some great recommendations, using cases were DNA ‘errors’ (by humans) had led to wrongful convictions, and yet DNA could exonerate. By making the system automatically generate alerts to ‘errors’ (both Type I and Type II) then we could save an awful lot of work from Innocence Projects around the world!
(To see a GREAT blog post on Type I & Type II erros in the justice system – read here… )
Of course, if mistakes are ‘discovered’, there needs to be sufficient transparency (and honesty/ ethical behaviour) among the State actors for that to be acted upon. In this article from Viriginia:
it is explained that the discovery that DNA matches with suspects to crime for whom someone was already sitting in jail, wrongfully convicted, has not led to any action: “Initially, the state Board of Forensic Science, which oversees the department, said only prosecutors and police were to learn the test results and that it would be up to authorities to decide the significance of the testing and to take any action.” Fortunately, legislators in Virginia have realised that this is not an acceptable state of affairs and made an amendment: “That the amendment got in the pending budget, Howell said, “is, I think, a reflection of an ongoing frustration with the process and the lack of transparency and the lack of effort made by the Department of Forensic Science.”
Such actions need to be replicated both across the US and internationally.
However, one issue in the UK at least, is that with our DNA database, crime scene samples are deleted from the database once a case is deemed to be ‘solved’ (i.e. there has been a conviction). It is rare for DNA profiles to be retained once someone is behind bars for a crime. This means that many people who may have been convicted based on the result of ‘errors’ cannot have that remedied unless there is access to original exhibits (and they still contain sufficient DNA to profile). This is a significant flaw in the UK system that so far, the government appears to be ignoring in it’s latest amendments to DNA legislation, despite the issue being brought to their attention (at the very least, in this report on “The Future of Forensic Bioinformation”).
See a related blog post I recently wrote on the subject here.