3 ms·

It's actually interesting to note that over time I've become increasingly skeptical of ferensics. My wife works daily in a genetics and neuroscience lab. She s

It's actually interesting to note that over time I've become increasingly skeptical of ferensics. My wife works daily in a genetics and neuroscience lab. She said typically one out of three experiments they do, even under the best conditions don't turn out. Given how little forensic DNA is often collected, and how often DNA either becomes contaminated or the experiment/process fails, I often question the validity of tests. My wife doubts the validity even more than I do lol

washadjeffmad 10y ago
And you're right to. Without proof of blind sample testing and rigorous QA on the process, there's no reason to trust the results. And that type of evidence, if accepted, would only wisely be used for ruling out suspects. Even considering the defense as the null hypothesis, the US legal process is not as scientific in some very important ways as it could be.
dogma1138 10y ago
Well it's not necessarily applicable. DNA sequencing in forensics isn't a full sequencing it's just a profile with an 1 in n match usually in the low millions which outside of direct same sex familial match is going to be highly unlikely to be a false positive. Forensic evidence doesn't exists in a vacuum it's looked at in context. It's also worthy to note that most cases aren't an episode of CSI and most convictions still rely on witness testimony and or confessions. Thats not to say that there haven't been bad practices in he past ranging from bad science to individual malpractice and even malice. But it's still better than just having he said she said, as forensics allow you to verify testimonies and build a plausible depicture of events.
- mikeash 10y ago
  Has the false positive rate been empirically evaluated? The fact that most convictions rely on witnesses doesn't make me any happier. Human memory is deeply flawed.

washadjeffmad 10y ago

And you're right to. Without proof of blind sample testing and rigorous QA on the process, there's no reason to trust the results. And that type of evidence, if accepted, would only wisely be used for ruling out suspects. Even considering the defense as the null hypothesis, the US legal process is not as scientific in some very important ways as it could be.

dogma1138 10y ago

Well it's not necessarily applicable. DNA sequencing in forensics isn't a full sequencing it's just a profile with an 1 in n match usually in the low millions which outside of direct same sex familial match is going to be highly unlikely to be a false positive. Forensic evidence doesn't exists in a vacuum it's looked at in context. It's also worthy to note that most cases aren't an episode of CSI and most convictions still rely on witness testimony and or confessions. Thats not to say that there haven't been bad practices in he past ranging from bad science to individual malpractice and even malice. But it's still better than just having he said she said, as forensics allow you to verify testimonies and build a plausible depicture of events.

mikeash 10y ago
Has the false positive rate been empirically evaluated? The fact that most convictions rely on witnesses doesn't make me any happier. Human memory is deeply flawed.