5 ms·

How can these fellows sleep at night? “Of 28 examiners with the FBI Laboratory’s microscopic hair comparison unit, 26 overstated forensic matches in ways that

How can these fellows sleep at night? “Of 28 examiners with the FBI Laboratory’s microscopic hair comparison unit, 26 overstated forensic matches in ways that favored prosecutors in more than 95 percent of the 268 trials reviewed so far.”

Amygaz 11y ago
It's like fsloth said earlier in this thread: someone needs to fall. Personally, twice I had FBI field agents as my neighbor, and I while my n=2 can hardly be generalized, my first impression (and you only get one chance to make a good first impression) is that the FBI doesn't get to hire the cream-of-the-crop. Maybe it's not the staff fault: the FBI seems to be investing a lot more in their PR department than in its departments actually doing the job in the trenches.
- dsr_ 11y ago
  The FBI cannot (or does not) pay as well for technical positions as nearly any private position. The background of agents (and, subsequently, management) is in either accounting or "criminal justice", neither one of which is a field noted for reassessment of theory or appraising scientific evidence. As a result, they tend to accept promises of accuracy that line up with their biases, and not dig too deeply into methods.
  - ams6110 11y ago
    This mindset is common in practitioners. Faith in the process, or the way it's always been done, or the way they were trained. Even in development, it's very common in corporate or government settings to find developers who know the one language and OS that they use, and the one set of patterns they use, and are not very curious about anything else. It's a job.
- rhino369 11y ago
  FBI gets to hire the cream of the crop for special agents but all other jobs are treated like dogshit. One example is that every non-special agent takes turns emptying garbage cans.
raddad 11y ago
The examiners need some stiff jail time. If their fake work puts a guy behind bars for 5 years then they should risk at least that same amount of time also. edited for coffeeless grammar
rhino369 11y ago
You are presuming they were maliciously making up evidence. The story isn't super clear but it seems like the examiners were overstanting certainty and that they were using dubious statistics. People do that sort of thing all the time, not out of maillice , but out of ignorance or bias. The article doesn't even talk about what the actual problems were. Was the forensics based on wrong assumptions or research later proven inaccurate. There is a difference between being wrong in good faith, being wrong because of bias, and straight up lying. Even the statistics in this article reek of bias. 95% of reports overstated the matches. By how much? Are we talking about 99.9% matches being called 100%. Or are we talking about 30% chance being called 100%. How many times was it substantially overstated? The biggest problem is that there isn't really an academic field for most of forensics. Which is pretty odd.
- Zigurd 11y ago
  Malice hardly ever feels like malice to the malicious. If the results are deeply malicious, why try to excuse them?
  - rhino369 11y ago
    Results aren't malicious. Malice is about intent. The reason you should care about the distinction is because many people think bad things only happen when people purposely act badly. But a lot of bad is done by people acting with good intentions or neutral intentions. This ironically leads people to falsely trusting systems to do the right thing. A bunch of well meaning FBI examiners can end up jailing innocent people if they are all using the same text book that got some facts wrong. Plus like I said, it's not even clear that innocent people were sent to jail because of this. At least it's not clear from the article, I haven't looked for the report that is mentioned.
    - Zigurd 11y ago
      Do you think lie detector operators are "well meaning" or are they knowing frauds?

Amygaz 11y ago

It's like fsloth said earlier in this thread: someone needs to fall. Personally, twice I had FBI field agents as my neighbor, and I while my n=2 can hardly be generalized, my first impression (and you only get one chance to make a good first impression) is that the FBI doesn't get to hire the cream-of-the-crop. Maybe it's not the staff fault: the FBI seems to be investing a lot more in their PR department than in its departments actually doing the job in the trenches.

dsr_ 11y ago
The FBI cannot (or does not) pay as well for technical positions as nearly any private position. The background of agents (and, subsequently, management) is in either accounting or "criminal justice", neither one of which is a field noted for reassessment of theory or appraising scientific evidence. As a result, they tend to accept promises of accuracy that line up with their biases, and not dig too deeply into methods.
- ams6110 11y ago
  This mindset is common in practitioners. Faith in the process, or the way it's always been done, or the way they were trained. Even in development, it's very common in corporate or government settings to find developers who know the one language and OS that they use, and the one set of patterns they use, and are not very curious about anything else. It's a job.
rhino369 11y ago
FBI gets to hire the cream of the crop for special agents but all other jobs are treated like dogshit. One example is that every non-special agent takes turns emptying garbage cans.

raddad 11y ago

The examiners need some stiff jail time. If their fake work puts a guy behind bars for 5 years then they should risk at least that same amount of time also. edited for coffeeless grammar

rhino369 11y ago

You are presuming they were maliciously making up evidence. The story isn't super clear but it seems like the examiners were overstanting certainty and that they were using dubious statistics. People do that sort of thing all the time, not out of maillice , but out of ignorance or bias. The article doesn't even talk about what the actual problems were. Was the forensics based on wrong assumptions or research later proven inaccurate. There is a difference between being wrong in good faith, being wrong because of bias, and straight up lying. Even the statistics in this article reek of bias. 95% of reports overstated the matches. By how much? Are we talking about 99.9% matches being called 100%. Or are we talking about 30% chance being called 100%. How many times was it substantially overstated? The biggest problem is that there isn't really an academic field for most of forensics. Which is pretty odd.

Zigurd 11y ago
Malice hardly ever feels like malice to the malicious. If the results are deeply malicious, why try to excuse them?
- rhino369 11y ago
  Results aren't malicious. Malice is about intent. The reason you should care about the distinction is because many people think bad things only happen when people purposely act badly. But a lot of bad is done by people acting with good intentions or neutral intentions. This ironically leads people to falsely trusting systems to do the right thing. A bunch of well meaning FBI examiners can end up jailing innocent people if they are all using the same text book that got some facts wrong. Plus like I said, it's not even clear that innocent people were sent to jail because of this. At least it's not clear from the article, I haven't looked for the report that is mentioned.
  - Zigurd 11y ago
    Do you think lie detector operators are "well meaning" or are they knowing frauds?