Fingerprint examiners claim officers lay on pressure
Bias may lurk in labs' ties to small police agencies
By Jim Stasiowski
Missouir Lawyers Media
A police officer rushes up to the fingerprint examiner and pleads for help.
The suspect sitting in the interrogation room is as guilty as can be, the officer insists, and a confirmed fingerprint match would surely bring about a confession.
Not exactly the environment to produce an unbiased examination.
But neither is it far-fetched in the sometimes boiling climate of crime-fighting. Police officers want arrests to stick, they want to get criminals off the street, and if the system allows them such access to the key processors of information, well, some officers will take advantage.
"They put pressure on you shamelessly," Pat Wertheim, a longtime fingerprint examiner, said. "I've felt it. "
But not at his present job, Wertheim said, nor would it happen at any lab accredited by the American Society of Crime Laboratory Directors. These days, he said, such pressure likely would occur in small police agencies.
Wertheim works for the Arizona Department of Public Safety, and he said that in his lab, police pressure is impossible. "We have an impervious firewall," Wertheim said. Police officers "can't get past the secretary. "
Wertheim, emphasizing that he is speaking for himself and not the agency he works for, recommends that defense attorneys find out what the examiner knew and when the examiner knew it.
"A good line of questioning [of an examiner] would be, 'Do the police have access to the lab? Do you listen to the police investigator before you do your exam? If the officer said he had an eyewitness, were you aware of that before you did the examination?'"
Bias potentially touches the fingerprint process in two ways. The first is the pressure police investigators apply before examiners start working. The second is in the actual examination process, which comprises four steps known by the acronym ACE-V.
The steps are: analysis, comparison, evaluation and, finally, verification. Fittingly, the last stands alone in the acronym, as it is the only time a second examiner plays a role.
After the first examiner declares a match, he or she turns the evidence over to a colleague.
David Grieve, an examiner retired from the Illinois State Police, said the verification has potential for bias because often the first examiner tells his or her conclusion to the verifying examiner.
"If you're giving it to the same person every time," he said, "a buddy system develops. That's a bad setup. "
Only a few agencies, such as the FBI, have enough examiners to spread around the verifications, and instead of telling the verifier the original conclusion, the first examiner may say, "I would like you to examine this," Grieve said.
When he is a verifier, Wertheim said, he relies on the scientific method: "I don't try to prove the other examiner right; I try to prove him or her wrong.
"I think most of them do that. That's good science. "
Itiel Dror, who is not a fingerprint examiner but has a Ph.D. in cognitive neuroscience from Harvard, is convinced bias taints forensic evidence.
Dror, who is affiliated with the University College London and works at Cognitive Consultants International, said his papers criticizing the judgments so-called experts make about forensic evidence, including fingerprints, have made him unpopular with many in law enforcement.
"The FBI has excommunicated me," he said, chuckling. "I'm one of their 10 most wanted. "
In 2005, Dror and two associates tested five experienced fingerprint examiners who had volunteered for the test.
Here's how the test went: At some point in the course of 2005, a colleague handed each of the five participants a latent and an exemplar. The exemplar, the colleague said, belonged to Brandon Mayfield, an Oregon attorney. The latent, the participants were told, was found at the site of the 2004 Madrid train bombings.
In that notorious case, the FBI said the latent belonged to Mayfield, but he was exonerated after Spanish authorities determined the print was another man's.
In reality, the latent and the exemplar had nothing to do with the Madrid case. Instead, each test participant had received a latent and an exemplar that he or she had identified, in 2000, as a match.
According to Dror's paper, only one of the five went against the suggestion of no-match and declared that the prints matched. Three of the five concluded the fingerprints were not matches, thus "contradict[ing] their previous decision," the paper said.
The fifth participant concluded the information was insufficient to lead to a firm conclusion.
Dror emphatically urges defense attorneys to cross-examine fingerprint examiners.
"Ask them, 'Are you aware of the vulnerability surrounding judgment and decision-making?'" Dror said. "If they say they are not aware, make the point that they're not keeping up with the latest research. "
But if they say they are aware of the dangers of bias, he said, "then ask, 'What are you doing about it?'"
Copyright 2009 Missouri Lawyers Media