Not so good news out of Florida

[mdavis]

Thanks, Pat. Well said, as always.

[L.J.Steele]

(1) I intentionally put the word "knows" in quotes, not to emphasize it as an underline, but as a conditional opinion. I think this is a justifiable position (your experience to the contrary) for several reasons. First, attorneys are as adept at discerning lies as police officers.

Oddly, the researchers tell us that neither police nor defense attorneys are all that good as lie detectors. There've been lots and lots of studies of different methods -- polygraph, statement analysis, eye movement, etc. -- most of it scores little better than guessing when actually tested. Those familiar problems of confirmation bias, tunnel vision, and suggestion affect investigators and counsel alike, sometimes causing us to miss or discount important information and come to wrong answers. Check out Saul Kassin's work on false confessions

http://www.williams.edu/Psychology/Facu ... ssions.htm

or

www.psychologyandlaw.com/false.htm

for some of the work in this area.

The really scary problem is when the client himself doesn't know what happened. Take a client who'se vulnerable to suggestion -- a child, someone with certain mental illnesses, prone to blackouts or memory loss - put them in a coercive situation with an authority figure (police detective) telling them they committed a crime -- and you can get internalized false confessions. The client comes to believe that he did something that, in fact, he didn't do.

These are also rare cases, but ones where investigators, the courts, and attorneys depend on the forensics folks to sort out the false confessions from the genuine ones. (And a good reason to remain skeptical if you hear that the suspect confessed before you analyze the evidences.)

When a client claims innocence, a defense attorney must consider that possibility, but also realizes that the majority are not innocent.

This is a very dangerous road for an attorney to go down. Yes, we all know the odds, but if the attorney starts out cynically, then those issues of confirmation bias, tunnel vision, etc., can lead to disaster. If one is lucky, then years later the client might get exonerated by future forensics. Can you imagine, for a sec., being one of the attorneys for the 200 or so Innocence Project exonerees who didn't believe the client when he said he didn't do it?

Second, a good defense attorney will carefully explain to a client that they must know the truth, that the truth is held in strictest confidence and not conveyed to the court, and that such information is critical in building the best defense possible. In other words, lie at your own expense.

Which doesn't mean that clients don't lie. Many don't trust their attorneys, especially public defenders, and assume that we're all in cahoots with the police and prosecutors to put them in jail.

Third, defense is given all pertenent evidence in the case through discovery.

This is assuming the prosecutors or police aren't playing "hide the ball" and are providing full discovery. Even if I grant complete disclosure in a case, from the defense PoV, police and prosecutors do push cases, especially if the crime is high profile, there's public pressure to convict somebody, the crime is particularly sensitive (DWI/OUI and child abuse cases come to mind), and/or the client is a pain in the neck or a frequent flier in the system.

[CT has appointed prosecutors, Mass uses elected ones -- in both cases, public pressure, tunnel vision, and the sheer momentum of a case can push all sorts of bad cases forward.]

Truely independent, reliable evidence is a good indicator. But that's what the fights between the defense bar and the prosecutors about forensic testing, eyewitness ID, false confession, and so on are all about -- getting that reliable evidence so the real bad guys get convicted of what they actually did, folks who are innocent (in whole or in part) aren't convicted, and we don't have exonerations a decade or two down the road when some new test finds our mistakes.

But the fingerprint community has a fantastic record of self-policing its own in the past.

Does it? Let me play devil's advocate for a second. Is there a central person or body who keeps track of mis-IDs and forgery/fraud claims? Who investigates them? Disciplines the person(s) involved? Any central authority to resolve disputed mis-IDs? I still hear disputed stories about whether Cowans was a mis-ID or a fabrication by dishonest examiners -- is there no one who can go to Boston PD, get the records, and publish the counterpart to the Stacey report so y'all can see what really happened?

I'm not saying whether or not there should be such a body, but reminding you of the threads talking about departments that don't train well and accept marginal work to clear cases. Is self-policing working there?

Again, before y'all get too worked up about Langill, let's see if the prosecutor files an interlocutory appeal (which I expect will happen soon if it is to happen) and/or tries to supply supplemental documentation. We haven't seen what was provided to the NH court -- it may be that what the judge wants is not complex and would be deemed feasible.

[sorbitol]

And things can get worse.


http://www.orlandosentinel.com/news/loc ... s-seminole

[mdavis]

Ouch! 25 years experience? Let's not pre-judge this examiner until other, respected examiners have seen the prints in question.

However, it does call to question things like vision impairment with advancing age, and over-confidence that I mentioned in an earlier post. Is this an accredited agency? Who is verifying the idents? Seems to suggest more than one problem with one examiner. Yet adverse publicity affects us all.

[Michele]

Here's the latest.

http://www.orlandosentinel.com/news/loc ... s-seminole

Here's my question, it looks like they are saying that the results of 3 of the 4 latents in question are really inconclusive. Shouldn't an agency have to state what the standard for a conclusion is before they can claim that there wasn't enough to identify a latent print?

[mdavis]

Well, how does one quanitfy/qualify the cutoff? Count "points?" It has always depended on the individual comparison. "I'd have to see the print" is a common answer to defense questions of thresholds.

Where are the supervisors at this lab???

[Michele]

If people can go back and look through cases and determine that something should have been called inconclusive then they must have some criteria to be able to do this. What is the criteria, and did Birks and McQuay know of this criteria (did the agency have it in writing)? I think we all have one (otherwise we wouldn’t classify prints from ‘awful’ to ‘not bad’ to ‘easy’ to ‘Helen Keller’s’) but we’ve never had to articulate what the criteria is.

In the past we’ve always said, “I’d have to see the print”. This agency now has 3 cases where they’ve said that 2 of their own certified examiners have crossed the line. How can you hold examiners accountable for crossing a line if you haven’t established where that line is? This may not be a scientific or legal issue, but I’d think it would be a good quality assurance measure for agencies to state their criteria.

It sounds to me like the courts are asking more and more what quality assurance measure we have in place to diminish erroneous conclusions (is having some sort of standard a good QA measure??). I’m not suggesting a numeric standard, although I do know of many agencies that have an unwritten numeric standard. I’m sure agencies have a lot of different quality assurance measures, I’m just wondering what some of them are. I’m also saying how important it is for agencies to have this kind of stuff in writing. We all recognize that we need these to back up what we did in court but it’s just as important to have these so we know what our agency expects out of us.

I’m not in management but I’m wondering if those that are in management recognize the need to state what is expected out of employees in case you ever want to hold them accountable? I could list several cases where examiners have taken the fall (practitioner error) when the agency’s lack of training and/or poor policies may have contributed just as much to the problem.

I’m also wondering what measures the examiners who are reviewing these cases are using to protect themselves against confirmation bias?? This is very similar to the Dror/Charlton study. I can’t remember who said this at the Boston Conference (maybe Bertram, Byrd, or Black) but someone mentioned that if you’re looking to confirm something then you’ll always find some sort of information to confirm your thoughts. Are the reviewers biased towards finding erroneous ID’s?? I guess only time will tell.

[mdavis]

Criteria. Used to be the old 12-"point" count. Then we strayed into the era of ACE-V and universally agreeing that there is no minimum number of "points" needed to effect an ident. Indeed, there is a well known example from (if I recall correctly) Illinois State Police in which a latent impression on a rifle scope had no "points" (as in Level II), yet was identified and verified by outside agencies. Even SWGFAST stays away from criteria.

If the ultimate decision is a subjective evaluation, agreed upon by two or more competent examiners, then the criteria must, by definition, be based on the individual ident. Have we not intentionally steered away from criteria? We have accepted the fact that we do not need to use quantitative measures ("points"). And we cannot easily describe qualitative criteria that will fit each and every latent print encountered. How do you articulate in a P&P manual the difference between "good" and "fair" and "not so good" and "lousy" and .... etc. I'm all for "criteria" if it will eliminate bogus idents, but isn't criteria an attempt to "fix what ain't broke" or "nail jelly on a tree?"

How about following the ACE-V standards? Did either of the examiners undergo proficiency testing? Apparently many idents were not verified, although apparently some were. Are verifications done at random, or on every ident? Are they truly being verified, or "rubber stamped?" Are these examiners recognizing the potential for bias that exists with all of us? Seems that might be the first line of defense after following industry procedures. As much as I hate the extra hassles of accreditation, this is one reason the courts are beginning to demand it of forensic labs. These are things everyone should be doing voluntarily.

We have one or two examiners who have declared idents where others disagree. Who are the others? Clearly the criteria between the groups differ. We don't know if the idents are valid or not. I think the issue is that there are too many examiners working too close to the "edge of the envelope", unwilling to accept the possibility of a "bad guy" getting away, and risking a career in the process. Fortunately, of the thousands of examiners working every day, the exceptions are quite rare however highly publicized they might be.

[L.J.Steele]

A defense attorney "knows" if his client is innocent or guilty.

The Innocence Project just announced its 201st exoneration. In the NY Times story, I noted this quote:

"Mr. Scheck noted that in about a quarter of the 201 wrongful convictions
that have been overturned with the use of DNA evidence, people had
confessed or admitted to crimes they did not commit. Mr. Halsey signed a
confession after 30 hours of interrogation, Mr. Scheck said. Mr. Halsey’s
lawyers said he had a sixth-grade education and severe learning
disabilities."

www.nytimes.com/pages/nyregion/?partner=rssnyt

Apparently Mr. Mr. Halsey had maintained his innocence from the beginning, according to his trial lawyers. What does the attorney "know" in this situation -- there's a confession to a brutal killing of two kids, and the client saying "but I didn't do it."

[mdavis]

Apparently Mr. Mr. Halsey had maintained his innocence from the beginning, according to his trial lawyers.

Mr. Halsey signed a confession after 30 hours of interrogation...

From the beginning to when? Sounds contradictory to me. What does DNA have to do with latent prints? Old cases that have been mis-tried (and no one argues that fact) should always be scrutinized if there is a reasonable question as to validity of evidence. But do we blame the court when they are handed a confession? Or do we blame the police for coercion? The case was corrected by the work of forensic analysts, not caused by them.

[L.J.Steele]

I merely comment on the proposition that defense counsel may know whether or not the client is factually guilty.

[sorbitol]

Holy Cow!

http://www.orlandosentinel.com/news/orl ... -headlines

[mdavis]

I would be very careful to assume the percentages of the "Innocence Project" to be representative of the whole of criminal cases. First, the source is obviously biased toward exoneration and would be expected to pass over "slam dunk" guilty cases in favor of digging into cases it would expect to see overturned. This would include intoxication, mental illness, consentual sex-turned bad, false reports and a host of other factors that are will known and documented in DNA type cases.

No argument here, just an extremely high level of caution against expectedly biased statistics. This is hardly a representative sample.

[L.J.Steele]

I would be very careful to assume the percentages of the "Innocence Project" to be representative of the whole of criminal cases. First, the source is obviously biased toward exoneration and would be expected to pass over "slam dunk" guilty cases in favor of digging into cases it would expect to see overturned. This would include intoxication, mental illness, consentual sex-turned bad, false reports and a host of other factors that are will known and documented in DNA type cases.

No argument here, just an extremely high level of caution against expectedly biased statistics. This is hardly a representative sample.

Agreed that it isn't a statistically valid sample set. One can ask IP for help with any case where there's DNA that might exonerate the suspect. (Partly, this is driven by the habeas rules that generally only allow one to bring in new evidence (unavailable at the time of trial) that tends to exonerate the defendant.)
So, no, they're not going to get involved in cases where the defendant admits the act but argues consent, or self-defense, or not guilty due to mental illness, or guilty of some lesser offense. Nor cases where no DNA evidence would likely be found -- typical street robberies, FREX.

What it does tell us are things like -- our witness ID procedures aren't working well, nor are the safeguards to keep flawed IDs out of court, or help juries assess them. Our procedures for handling false confessions are similarly weak. Ditto problems with "snitch" testimony. Cowans told us that there was a problem with fingerprint IDs in Boston -- until and unless Boston publically produces the data we won't know what went wrong there.

[Michele]

Wow, they handle things quick in Florida!

http://www.orlandosentinel.com/orl-prin ... -headlines