4 ms·
> Black explained her rationale, but conceded that she didn't have statistics showing the likelihood of the hands matching. It looks like an open & shut case t
by roenxi 5y ago
> Black explained her rationale, but conceded that she didn't have statistics showing the likelihood of the hands matching.
It looks like an open & shut case to me, but this sort of thing should be grounds for concern in a court case. Evidence that can be statistically quantified should be. Especially novel identification techniques.
I'm glad that more research went in to the technique before the convictions started to happen.
- nyolfen 5y agootoh 'more research' seems to be a database of 1000 individuals compiled and analyzed by an individual. this appears to be founded on ad hoc visual inspections without theoretical underpinnings, and seems like a scientific cargo cult like many other forensic science methods.
- ttyprintk 5y agoAgreed. Before admitting it in court, it should reach a confidence of disproving someone’s identity before being believed to prove* another identity. Edit: (*) beyond a reasonable doubt, as expressed in the law.
- mcphage 5y ago> Before admitting it in court, it should reach a confidence of disproving someone’s identity before being believed to prove* another identity. Which is exactly how it was used in the case in question—Black said it couldn’t prove the identity of the attacker, but was enough to show they couldn’t exclude the girl’s father.
- LorenPechtel 5y agoI don't think her case needed more research. She said her father was molesting her. She set a webcam, somebody was certainly molesting her. Why would one think she was identifying the wrong person?