FBI admits errors in DNA match calculations

[pbray]

In the latest news from the FBI crime labs, an admission that they've been overstating the likelihood of DNA matches, a problem that dates back to at least 1999.

This is part of a series of reviews that have revealed systemic problems with forensic evidence provided by the FBI. Back in 2012 they revealed similar issues with hair & fiber evidence, these dating back to the 1980s. Not to mention the earlier well-publicized problems with gunfire residue evidence and fingerprint analysis.

Back in 2009 the National Academy of Sciences went before Congress with a scathing indictment of the state of forensic science at all levels of law enforcement in the US, pointing out that the lack of standards and scientific rigor meant that junk science was routinely being presented, unchallenged, at criminal trials. In an optimistic view, one could say the latest report from the FBI shows that they've committed themselves to reform, and are being open about the past.

Or you could say that they are just as likely to learn from this as they learned from the FBI lab scandals in the 1990s...which is to say, not at all. Bad FBI science.

Comments

[mtlawson.livejournal.com]

The snarky part of me wonders just how many of those old Forensics Files episodes will need to be revised.

The not-so-snarky part of me wonders just how many criminal convictions will be overturned.