CLPEX Chat Board

thought this was interesting.

http://www.ajc.com/metro/content/metro/ ... _0507.html

Some people shouldn’t be allowed to talk to the media.

GBI spokesman John Bankhead said:
“The print was similar to Presnell's.” (How does he know? It sounds like the ‘amazing amount of similarity’ statement being used again. I wonder where he heard that
and
"The problem, when they come up on screen, there's nothing to ID who they belong to," (That doesn’t seem to be a problem for the rest of the world)
and
"They were a match, but not the suspect's prints." (so it wasn’t an erroneous ID?)

He must have been hired for his ability to put a positive spin on things.

The DA says, (Drew Lane):
“the error could have been worse” (as one of my coworkers said, he obviously hasn’t spent 2.5 years in jail)
And
“the setback doesn't automatically make Presnell innocent” (SETBACK? That’s an interesting choice of words)


I’m wondering about this statement”
“Double checking his work later, the examiner found that……”
Do most agencies arrest someone after the ACE phase and verify years later, right before the trial?

Quite possible that the "double checking" was actually preparing for trial and getting that old sinking feeling when the defendant's numbered digit was obviously not something that would have been matched to the latent. For cops, the letter begins, "Dear Chief: No one was more surprised than I..."

Okay, I'm still not clear on what happened here. Probably a bit the wretched reporting and a bit terminology. When they say, "...pulled up on the screen..." do they mean a fingerprint card or a single digit? We have a Cogent AFIS system that does not print out the name of the person on the exemplar when it prints out. Conceivably, he could have printed out an exemplar and gotten mixed up as to who it belonged to...that happens occasionally in our office.

When they say, "...in checking his work later..." how MUCH later? Did no one do verification at the time? Did they just assume the exemplar was the correct one?

I also had a case where palmprints entered into the Cogent database, (by numerical identifier), were for the wrong person. I used them to mistakenly eliminate a suspect. I found out later, purely by serendipity, that the palmprint cards I had used were wrong; got the correct prints and made the ID! Scary.

I HATE LIVESCAN!

Luckily, I am the Captain Stubing of navigating negative PR. This mishap was obviously taken out of context by a vast left wing conspiracy headed up by Masons, the Illuminati, Skull and Bones, The ACLU, and Big Oil. What happened at the GBI was merely a "clerical error"...

Say it with me people, "it was a clerical error"

Sincerely,

Boyd Baumgartner
Fingerprint Apologist

Seems like prima facie evidence in support of requiring a hard-copy of exemplar prints complete with demographic info prior to confirming an identification (individualization).

For a variety of reasons (including elimination of bias, privacy & security concerns, and to facilitate computer cross-indexing in relational databases), most AFIS systems have provided on-screen fingerprint images for comparison without benefit of a name attached, accessible only through cross-indexing via a separate search through a different database.

Based on the limited info available regarding this case, it appears to offer a prime example of why we should be very cautious about going to a 'paperless' environment, attempting to make identifications on-screen, as opposed to only making final decisions with a hard-copy. Of course, I have also seen traditional inked fingerprints in years past with the wrong name... both intentional & accidental. The difference in this case being the error would likely have been preventable with a hard-copy policy in place.

Without knowing more, we have no idea if this involved a "paperless" environment. Since the only description is that the "examiner mistakenly pulled a digital image onto his computer of a print from Wheeler's daughter who lived at the house..." we don't know if he had a hard copy of elimination prints and they were scanned, or that the elimination prints were the result of using a Livescan device to put them in a database, or that the daughter already had them on file.

It could be that the examiner grabbed the card thought to be the suspect's card and scanned only the finger (or fingers) into a digital imaging program for the comparison. But the elimination card was the one that was scanned, and there wasn't a mechanism built into the procedure to catch this later (peer review in the verification stage). If there was, then an entirely different can of worms has been opened.

Here's another article with a little more information regarding the digital system used (assuming it's correct):

http://wsbradio.com/news/050608falsearrest.html

A fingerprint lifted from the scene was to be compared to the suspect's inked prints as well as to the elimination prints from the victim's daughter and other family members who lived at the residence.

The examiner used a new digital imaging system to record and store the prints for later comparison on his computer.

When he opened the print images for comparison, he mistakenly opened the elimination prints from the daughter instead of the suspect's prints, which led to the error.

So there may have been hard copies for elimination, but it sounds as if they were scanned into this system for storage.

Steve Everist wrote:Without knowing more, we have no idea if this involved a "paperless" environment.

True enough, as we are all speculating on various scenarios based on limited info. My point remains, however, that this type of error could happen in a 'paperless' environment with current AFIS and other imaging system configurations. Whether through the use of hard-copy confirmation of ID's, or other means such as flagging each print image with the subject's name, the chance of error may be reduced.

Dennis Degler wrote:Unfortunately, this won't ensure that mistaken identity will never happen again, for we still occasionally place incorrect demographic info on hard copies and get the arrestee to sign the wrong cards.... I'm not saying we shouldn't take such mistakes seriously, but that we shouldn't be suprised [sic], nor overreact when they happen.

I did offer a similar caveat in my post regarding incorrect demographic info, and I agree on the caution about overreaction. However, we should also be wary of too easily dismissing or turning our backs on lessons to be learned.
Regards, RLT

Couldn't this have been avoided by comparing the actual latent (whether it be a card or photo) to the actual tenprint card?

Where do you write and sign that the latent is an ID on a screen comparison?

Whatever the circumstances, no excuse in my book.

From reading both articles sounds like the exemplars were scanned in and worked from there.

In my agency we have to keep the original card or a copy of it that we compared because of accreditation even if it comes out of an AFIS Digital Data Base (because of substitution) so we have been scanning them in after examination so the copies can be destroyed (shredded). But that is after the examination and before the review.

Now I know that shredding part caused a few of you to sit bolt upright. Shredding cards you have made ID's on----Are you NUTS. Well probably but that is besides the point.

See our cards are digital so we print them out and do our comparison, mark on the cards with the information we require. Lab number, examiners initials, etc and then scan them IF they are of the victim or an officer or someone not the suspect. Once scanned back into LIMS we can shred them instead of trying to find a place to store them.

Maybe GBI was doing something similiar but was scanning them in first before comparison. Or as Gerald said, perhaps a quick look before court caused it to surface.

Betcha some things change though.

If I can just do 4 more years without a mistake or error then I am home free, and I do not want to hear any cute remarks from those already on retirement out there. I am hurrying as fast as I can to get there.