Merged FBI gave flawed / overstated testimony in nearly all trials before 2000

[uvar]

No specific trial, but potentially one hell of an error.

http://www.washingtonpost.com/local...515-11e4-b510-962fcfabc310_story.html?hpid=z1

The Justice Department and FBI have formally acknowledged that nearly every examiner in an elite FBI forensic unit gave flawed testimony in almost all trials in which they offered evidence against criminal defendants over more than a two-decade period before 2000...

The cases include those of 32 defendants sentenced to death. Of those, 14 have been executed or died in prison...

The FBI errors alone do not mean there was not other evidence of a convict’s guilt...

 

[Charlie Wilkes]

Fascinating. Junk science, including bogus hair analysis, is a huge factor in wrongful convictions. Essentially there are no meaningful standards. We're presently looking at the Mark Lundy case, where a dubious forensic test was the key evidence used to convict someone who is almost certainly innocent.

 

[rjh01]

How many times are their effective control groups in such testing? By controls I mean as in a scientific experiment.

 

[Desert Fox]

I have often heard argued that the idea of "Convict at all costs" is an anomaly but every time the situation is looked under the hood, it seems more systematic than anything.

 

[luchog]

I have often heard argued that the idea of "Convict at all costs" is an anomaly but every time the situation is looked under the hood, it seems more systematic than anything.

I don't know that it's necessarily "convict at all costs", as much as it is an artifact of the forensics labs being under the auspices of law enforcement, so there is a mindset of finding evidence to convict, rather than following where the evidence points with no preconceptions or expectations. Combined with the fact that having the labs so closely associate with law enforcement and government means that they are going to have access to resources and funding that the vast majority of defendants won't.

The obvious fix for the problem would be to divorce the forensics labs from law enforcement, or at the very least, submit all evidence for review to a properly-certified commercial third-party lab agreed upon by the defense.

 

[Desert Fox]

I don't know that it's necessarily "convict at all costs", as much as it is an artifact of the forensics labs being under the auspices of law enforcement, so there is a mindset of finding evidence to convict, rather than following where the evidence points with no preconceptions or expectations. Combined with the fact that having the labs so closely associate with law enforcement and government means that they are going to have access to resources and funding that the vast majority of defendants won't.

The obvious fix for the problem would be to divorce the forensics labs from law enforcement, or at the very least, submit all evidence for review to a properly-certified commercial third-party lab agreed upon by the defense.

Some double blinding would be really nice as well. Also have one person develop the profiles and another look at them.

I still there is a huge problem with people being convicted on what is basically garbage evidence.

 

[Samson]

Some double blinding would be really nice as well. Also have one person develop the profiles and another look at them.

I still there is a huge problem with people being convicted on what is basically garbage evidence.

How about one team independent of the police do the crime scene and victims, and analyse in one lab. Another team takes the suspect samples and analyse in a different lab. A third team of highest level scientists determines the relevance of the samples and matches. Only what the third team reaches concensus on is returned to the police and prosecution.

 

[Desert Fox]

How about one team independent of the police do the crime scene and victims, and analyse in one lab. Another team takes the suspect samples and analyse in a different lab. A third team of highest level scientists determines the relevance of the samples and matches. Only what the third team reaches concensus on is returned to the police and prosecution.

That seems impractical at that level.

 

[Samson]

That seems impractical at that level.

Expensive, less evidence, fewer prosecutions, fewer convictions.

 

[Desert Fox]

Expensive, less evidence, fewer prosecutions, fewer convictions.

We do need to try something like a happy medium. . . . .There are so many other problems such as public defenders not even doing cursory investigations.

 

[rjh01]

How about one team independent of the police do the crime scene and victims, and analyse in one lab. Another team takes the suspect samples and analyse in a different lab. A third team of highest level scientists determines the relevance of the samples and matches. Only what the third team reaches concensus on is returned to the police and prosecution.

Expensive, less evidence, fewer prosecutions, fewer convictions.

We do need to try something like a happy medium. . . . .There are so many other problems such as public defenders not even doing cursory investigations.

I think Samson is onto something here. How about having an exercise where test samples are sent to the lab? The difference here is that the results are already known but not to the people who are doing the testing. If they give the wrong results then the lab is in serious trouble.

 

[Elaedith]

The same applies to any type of forensic analysis involving pattern matching, for example fingerprints:

http://www.cci-hq.com/Dror_FSI_contextual_influences.pdf

and even some types of DNA evidence :

http://www.cognitiveconsultantsinte...ty_and_bias_in_DNA_mixture_interpretation.pdf

Blinding isn't considered necessary because people fail to understand the unconscious nature of biases, believe that they can determine absence of biases through introspection, and that 'expertise' makes them invulnerable to bias. Aside from that there are other reasons for lack of motive to use procedures that be perceived as reducing the chance of convictions.

 

[Samson]

Fascinating. Junk science, including bogus hair analysis, is a huge factor in wrongful convictions. Essentially there are no meaningful standards. We're presently looking at the Mark Lundy case, where a dubious forensic test was the key evidence used to convict someone who is almost certainly innocent.

I think Samson is onto something here. How about having an exercise where test samples are sent to the lab? The difference here is that the results are already known but not to the people who are doing the testing. If they give the wrong results then the lab is in serious trouble.

I am curious about another matter.
Is America less safe after the acquittals of Casey Antony and OJ Simpson?

 

[luchog]

I think Samson is onto something here. How about having an exercise where test samples are sent to the lab? The difference here is that the results are already known but not to the people who are doing the testing. If they give the wrong results then the lab is in serious trouble.

I think that type of testing is already part of the certification process for private, commercial labs. Not sure if there is any certification process for labs under the auspices of law enforcement.

 

[luchog]

I am curious about another matter.
Is America less safe after the acquittals of Casey Antony and OJ Simpson?

Those cases did not signal a change in the validity of the processes involving evidence, or any other flaws in the criminal investigation or trial processes. What they did, was point out the glaring flaws that already existed, but which were either little known or actively hidden. It made clear what so many people who have been victims of the system have always known. That level of incompetence and manipulation was there all along.

 

[Chris_Halkides]

extraneous information and the expected result

The obvious fix for the problem would be to divorce the forensics labs from law enforcement, or at the very least, submit all evidence for review to a properly-certified commercial third-party lab agreed upon by the defense.

IIRC this was a suggestion made in 2009 by the National Research Council in its report on strengthening forensic science, and I agree that it is a good idea.

In the essay "CSI for real" Roger Koppl wrote, "Larry S. Miller demonstrates an excellent example of cognitive bias at play. He asked a group of 14 students trained in hair analysis, all of whom met the basic requirements for expert testimony on human hair identification in courts of law, to examine four cases each. For each student, two cases were presented the usual way: They were given two samples and told that one was from the crime scene and the other from the suspect. The other two cases were presented through a forensic lineup. The known sample from the imaginary crime scene was compared to five suspect-known hair samples. In all 56 cases, there were in reality no true matches. The first group of cases yielded an error rate of 30.8 percent; the second group an error rate of only 3.8 percent.¹² Miller’s study illustrates how evidence is often presented to the forensic scientist in a bias-inducing manner. The samples are labeled as coming from the defendant or from the victim and are frequently accompanied by a synopsis of the investigation indicating the reasons that the investigators believe the suspect is guilty. This protocol cues the forensic worker to the expected or correct result."

 

[marplots]

The lesson here is to distrust science. A good reminder for skeptics, even when we aren't on trial.

 

[Desert Fox]

The lesson here is to distrust science. A good reminder for skeptics, even when we aren't on trial.

I would argue that it means that one should use the tools that science has created to reduce bias.

 

[Trebuchet]

The lesson here is to distrust science. A good reminder for skeptics, even when we aren't on trial.

I think a better lesson is to distrust law enforcement.

CNN finally heard about this this morning and his been talking about "flawed" testimony all day. A better word might be "false" or even "perjured".

 

[Beerina]

I thought things went through a vetting process, where the first cases to allow them went into the science in detail during the trial itself.

Only later is it accepted at trials with much less detail into the justification.

DNA was good. Polygraph was bad.

Bullet barrel scratches good, then bad. Fingerprints good, then kind of good. Hair good, then bad.

Apparently nobody actually does a statistical analysis to design a partitioning algorithm followed by double-blind tests of that algorithm.