Is this lie detector telling the truth?

By Christina Lewis

Richard Allen Nicolas' story seemed suspicious. An unknown gunman shot at his car, killing his 2-year-old daughter, Aja. Plus, police found that his muffler was cold, although Nicolas said he had been parked a short time.

But police had little hard evidence against Nicolas until he was hooked up to a computer voice stress analyzer (CVSA), a machine designed to detect lies by monitoring small vibrations in a person's voice, and failed the test. He confessed.

Nicolas is now serving life in prison without parole for his daughter's murder.

To hear some police officers tell it, the CVSA, a laptop device designed to detect lies by monitoring a person's voice, is the greatest thing since handcuffs.

Imagine: no need to corner a suspect, attach multiple wires to him and have an expensive professional ask him a number of specific control questions. To use the voice stress analyzer, police just play a tape of the suspect talking, run it through the $10,000 laptop, and find out if he's truthful. So easy to use, the CVSA can surreptitiously verify the honesty of prospective employees, ferret out fraudulent insurance claims, and resolve disputes when it's one person's word against another's. Or so its creators say.

According to the National Institute for Truth Verification, the Palm Beach-based company that makes and markets the CVSA, you simply attach a microphone to the subject (or to the tape recorder or phone line), run the sounds into the device and get your results.

Graphs on the voice stress analyzer allegedly indicate whether a suspect is lying or telling the truth.

The device monitors the frequency of "micro-tremors," vibrations in the voice undetectable to human ears, that increase when a person lies. While the subject is speaking, the CVSA measures and displays any changes in the vibrations.

For each voice pattern the machine shows a graph: a high peak denotes a true statement, while a jagged plateau indicates a lie.

Approximately 1,200 law enforcement agencies use the CVSA, and supporters say that its convenience and accuracy will lead more departments to let their polygraphs, the more well-known truth verifier, start collecting dust.

But not everyone is enthusiastic about it. In at least a few high-profile cases, the device has appeared to be wrong. And a number of lawyers, civilians and scientists say the CVSA has no scientific validity.

"It's basically a Ouija board," said Nevada lawyer Ian Christopherson, who successfully defended a juvenile probation officer against a rape charge prosecuted largely on the basis of a voice stress test.

Christopherson's client, Vincent Sedgewick, took a CVSA test assuming it would clear him. Instead, it pointed to him as the rapist. When DNA testing did not match him to the sample recovered from the victim, however, police arrested him as an accessory based on the CVSA results. Although CVSA tests cannot be used as evidence in court, in some states it can be used to support an indictment.

Critics of the CVSA say that officers' belief in the infallibility of the lie detector is one of its greatest dangers.

In the case of 14-year-old Michael Crowe, a California teen who confessed to his older sister's murder, investigators administered a CVSA that allegedly showed Crowe was lying when he claimed to know nothing about his sister's murder. Although there was no physical evidence linking Michael to the crime, investigators continued interrogating him after his voice stress test and got him to confess. A judge later threw out his confession, and another man was charged with the crime.

"If you take a look at the Crowe case you have police relying upon this quackery directing their investigation in the wrong direction," Christopherson said. "That's why it's incredible law enforcement allows this out there because people start believing that this device can actually be accurate."

A study of the CVSA conducted for the Department of Defense in the mid-1990s concluded that the device detected deception at rates no better than chance.

In one of those studies the polygraph performed just as poorly, but Dr. Victor Cestaro, the researcher who conducted the tests, noted that a wealth of other studies have shown the polygraph's results to me more reliable.

"I'm not saying [the CVSA] doesn't have possibilities," Cestaro says. "I just fail to see where there is any scientific support of it."

The National Institute for Truth Verification claims, however, that its product has a much higher accuracy rate, based on the positive responses of police officers who have used it.

Capt. David Hughes, the company's executive director and a former chief of detectives for West Palm Beach, Fla., says he regularly gets phone calls from police officers thrilled with the device and ordering more.

"If someone fails the test and confesses then it must have worked," Hughes said. "No one is calling back and saying that they're having all these failures."

Lt. Rick Miller of the Vestavia Hills, Ala., Police Department has been using the device for eight or nine years and sees no reason to stop. "After I run an exam I have a much better understanding if this is someone I need to focus my attention on or not. And there's never been anything thus far to indicate that my [CVSA] analysis of a suspect has been wrong," Miller said.

Does accuracy matter?

Some argue that in the grim game of catching criminals, the CVSA gets the only results that matter. In other words, just because the lie detector isn't always accurate doesn't mean the confession isn't real.

Ever since the Supreme Court outlawed the torture of suspected criminals, police have been using lie-detectors as a tool in getting suspects to confess. Some early models were made by officers themselves with no pretense toward real truth-telling.

The most infamous is the now-improbable photocopier scam in which a police officer typed the word "False" on a piece of paper and placed it in a copying machine. When the suspect denied his crime, the cop hit the copy button.

Jeffrey Nance, a former undercover cop and author of "Conquering Deception," which examines different ways of detecting lies in everyday life, prefers to bypass the question of scientific accuracy and just think of lie detectors as interrogation tools.

"I don't think there's a foolproof way to tell if a person is lying or not," Nance told Courttv.com. "There may be people who don't like me saying that but it's the truth."

"It's the fear of the machine that gives it its greatest power," he said. "Even before they get on the machine they may give up information just because [they are afraid] the machine will show that they've lied. And nobody wants to be a liar."