With all of the snappy glamour of a noir film, polygraphs pose an attractive solution in the public eye: Just hook somebody up to a machine to figure out whether he or she is telling the truth.

The exams, which experts say aren’t well-understood, have been subject to controversy over the past decade. Their scientific validity has been called into question, and many states still don’t allow polygraph exam results into the courtroom.

While polygraph exam results aren’t admissible in trial in Oregon, the exams have played supporting roles in the investigation of several notable Bend cases. Not least of these is that of Daniel Norquist, who was released earlier this month after the district attorney dismissed a murder charge against him in the July 25 shooting death of his landlord and neighbor, Andrew Cordes.

John Meyer, a suspect in the 2011 disappearance and death of his wife, Sandy Meyer, killed himself the day after failing a polygraph test at the Bend Police Department, according to Bulletin archives. Police found other evidence before and after his death, including notes, research and emails that made him the prime suspect.

A clean polygraph test eliminated a locksmith as a suspect of a burglary in which approximately $2 million worth of gold, silver and jewelry was stolen from a Northwest Bend home in 2008. In July 2012, a 26-year-old Bend man was arrested after failing a polygraph test during a child abuse investigation, according to Bulletin archives.

Norquist was released from jail after Deschutes County District Attorney John Hummel moved to dismiss the murder case against the 34-year-old Bend man for Cordes’ death.

Hummel said the polygraph exam was one component among other evidence he considered in deciding to dismiss the case.

The district attorney said Thursday he decided to request that Norquist take a polygraph exam before continuing the second week of grand jury proceedings in the case. In the days after Cordes’ death, Norquist had undergone a lengthy interview and a crime scene walk-through with police. He had presented a consistent, detailed account of the night he killed Cordes, Hummel said.

“In a weird way, the polygraph didn’t add anything to this case,” Hummel said. “It would have been interesting if it had contradicted the other evidence at that point, but we didn’t get there.”

Public records of the polygraph examination show the examiner, Chris Kaber, a former state trooper who now runs a polygraph services firm out of Klamath Falls, found “no deception indicated” in answers to “relevant questions which were claiming truthfulness to written statements.”

Kaber declined to comment on the exam after speaking to Norquist’s attorney, Jonathan Char.

“It is his client’s desire that I not discuss the results with anyone he hasn’t ‘released’ me to speak to,” Kaber wrote in an email Thursday. He did not respond to a request for comment on the use of polygraphs as investigative tools.

The statements Norquist made when undergoing the polygraph examination primarily concerned the minutes when Cordes allegedly threatened him and remained in his apartment despite Norquist’s entreaties for him to leave.

Norquist wrote down and spoke aloud the following statements, according to Kaber’s report:

• “Andrew began walking down the hallway, it was at this point I told him he needed to leave.”

• “As Andrew began walking up the stairs, I told him not to come up ther (sic) stairs. I said this more than once.”

• “Multiple times during the night andrew (sic) would tell me that he wants to kill me. He also put his hands around my neck several times.”

Hummel said the polygraph focused on a specific aspect of Norquist’s claim that he killed Cordes in self-defense. In Oregon, to justify the use of lethal force against someone else, one must reasonably believe that he or she is in imminent, life-threatening danger, or that the other person is committing or attempting to commit a felony involving the use or threatened use of force against someone or burglarizing or attempting to burglarize a home.

The polygraph was not a means to determine if Norquist reasonably acted in self-defense, but a way to check the truthfulness of Norquist’s assertion that Cordes remained in his house despite being told to leave. In that case, Hummel said, Cordes would have been committing burglary.

“Is (the polygraph) appropriate as a tool to aid me in determining whether the use of force was justified? No, it’s not,” Hummel said. He later added: “We were really interested in testing our belief that a burglary had been committed. I have said, and I continue to say, I think Norquist should not have shot this guy … (but) under Oregon law he would have been allowed to if burglary was being committed, and the person who committed burglary had an intent to commit another crime.”

Reliability

Proponents of polygraphs say the tests may be more reliable than the abilities of humans alone to discern deception and that the test has a basis in science — humans have measurable physiological reactions to stimuli.

Moreover, the tests measure neither falsehood nor fact, but differences in physiological responses to test questions, and in turn their results are not definitive but a probability of error, according to Raymond Nelson, a psychologist and the president of the American Polygraph Association.

“When people are telling the truth, we’re going to see a certain pattern of responses,” Nelson said. “When people aren’t telling the truth, we’re going to see a certain pattern that we can quantify.”

In 2003, a committee of scientists assembled to investigate the validity of the polygraph exam and authored a report called “Polygraph and Lie Detection” published by the National Academy of Sciences. Professor Steven Fienberg, the statistician who chaired the study and is a professor of statistics and social science at Carnegie Mellon University, maintains those physiological reactions can be attributed to many different factors, such as fear or anxiety.

“Basically what you tend to see there is that what polygraph exams are doing are trying to measure signals from the … nervous system from very noisy channels that also measure responses to other things, like fear,” Fienberg said in a phone interview Thursday. “Any time you use them even in concert with one another, you make errors. The question is how much error there is going to be. And can you quantify it?”

He and another scholar on the committee, Professor David Faigman of the University of California, Hastings, said in separate interviews last week that investigators should tread carefully in using polygraphs.

Nelson countered that the probability of error can be quantified, and probability itself is key to what the test does.

Fienberg said that at the time of the 2003 National Academy of Sciences study, the studies collecting data on polygraph tests then weren’t uniform or rigorous.

“If you look at any of the scientific literature, there aren’t a lot of settings where people use exactly the same series of questions set up in the same kind of way to allow you to systematically assess things,” Fienberg said. “The problem I personally have in trying to come to grips with their use in criminal settings … is that basically somebody’s making up a set of questions on the fly. There’s no way to assess whether or not you would get a similar result or with any kind of assessment of error if you did it again because it’s unique to that individual.”

But Fienberg said that it would not be “unreasonable” to use the results of a polygraph exam in concert with other evidence in a criminal investigation.

Nelson, the American Polygraph Association president, conceded that the test wasn’t perfect — but if it was a certain determinant of credibility, it wouldn’t be a test. The profession has also taken steps to standardize testing procedure and minimize the amount of interference with the test results, he said.

“I think there’s a lot of frustration around the polygraph because it’s not perfect,” Nelson said. “But what people neglect when they say that is (that) no test is perfect.”

“The ethical role of tests is not to make decisions,” Nelson continued. “It’s to give information.”

The tests in trials

Carrie Leonetti, an associate professor of law at the University of Oregon law school, said in an interview earlier this month that one of the objections to allowing polygraph results into the courtroom has to do with the role of the judge and jury.

Both are considered “triers of fact,” and are charged with the task of assessing a witness’s credibility, Leonetti said. Admitting polygraph exam results would usurp that responsibility from the judge and jury.

Faigman, the UC-Hastings professor, said that although the credibility assessment is a factor in polygraphs’ exclusion from trial, some courts have excluded polygraphs in part because their results are unreliable.

“It’s hard to know how the polygraph was administered,” Faigman said. “All polygraphs have validity and reliability problems, but some are worse than others. The general conclusion that we reached (in the 2003 study) was that polygraphs are probably better than flipping a coin. They may be better than an average person. There are no good comparative tests of whether people are better lie detectors than a machine.”

And, Faigman said, the very condition of taking a test that the subject perceives as foolproof may skew the results.

“A lot of police departments and a lot of agencies use the polygraph, not because they think it’s valid, but because they think it’s useful,” said Faigman. “It’s actually a very good prop in an interrogation.”

Bend Police Detective Pat Hartley, who has been conducting polygraph exams for police, private investigators and attorneys for about two years, said he intends to administer tests without an agenda — he doesn’t conduct the investigations, just the exams. While there isn’t consensus in the scientific community, the test can add “incremental validity” to a suspect’s account of what happened, he said.

“It’s going to allow decision makers to get more information to make a more probabilistic decision,” Hartley said Friday. “There is no such thing as Pinocchio’s nose. That would be fantastic.”

— Reporter: 541-383-0376,

cwithycombe@bendbulletin.com