The Fallibility of Fingerprints

Thanks Alexander


Fingerprint evidence unreliable(DNA next ?).

January 29, 2004

Fingerprint science may be unreliable By Mark Henderson, Science Correspondent

FINGERPRINT evidence is much more unreliable than the public, police and courts think and rests on foundations that have never been rigorously tested, an investigation by New Scientist magazine has suggested.

The idea that two identical fingerprints must come from the same person is based on research that has never been peer-reviewed by independent experts — the gold standard for scientific proof.

Although there is no suggestion that fingerprinting is not a valuable forensic tool, many scientists are concerned that it is often seen as infallible. In fact, the technique is so old that its reliability has never been properly assessed with modern statistical methods, and its error rate is not known.

The only modern study to address the issue has never been published in full, and its finding that the probability of a false match is virtually zero was based on a sample of just 50,000 prints.

“Contrary to what is generally thought, there is little scientific basis for assuming that any two supposedly identical fingerprints unequivocally come from the same person,” New Scientist said.

The magazine’s concerns stem from scientists’ criticisms of a US study conducted in 2000 by the FBI and the company Lockheed Martin. The research was commissioned by the US Justice Department after a court case revealed that fingerprints failed to meet a key criterion for forensic science evidence — that the error rate is known.

The study purported to find an error rate of 1 in 1097— 10 followed by 96 zeros. As only 1011 human fingerprints are thought to have existed, this error rate is essentially zero.

Only a summary of the findings, however, have been formally published, and many experts believe that the methodology was flawed. Particular concern surrounds the sample size: although the study made 2.5 billion comparisons between prints, only 50,000 actual prints were used.

James Wayman, of the US National Biometric Centre at San Josť State University, said: “The (US) Government is comfortable with predicting the fingerprints of the entire history and future of mankind from a sample of just 50,000 images, which could have come from as few as 5,000 people.” The 1097 figure was an “absurd guess”, he said.

Further doubt has been cast on the Lockheed Martin research by a study published last month in the International Statistical Review by David Kaye of Arizona State University. It showed that the Lockheed investigators found three instances in which supposedly different fingerprints looked unusually alike. Although these images were excluded from the study as mistakes, they should have been included to show how fallible comparisons can be.

The critics of fingerprinting say that science can rarely prove error-free absolutes, and forensic scientists should not behave as if it can.

Jim Fraser, of the UK Forensic Science Society, said: “The idea that there is something about fingerprints that is fundamentally different from any other area of human knowledge concerns me. There have to be errors.”



Please contribute today - buy our books - and spread the word to all your friends!

* * * Back to the Home Page of John "Birdman" Bryant, the World's Most Controversial Author * * *