After Chicago police repeatedly questioned Donny McGee about the murder of his elderly neighbor, a detective on the case asked him to "face the truth" and take a polygraph.
McGee was taken to the polygraph room but was never given the test.
Instead, polygraph examiner Robert Bartik later testified, McGee stood up as Bartik opened the door and began confessing — even before the examiner could say a word.
McGee denied he confessed. When the case went to trial, a jury found him not guilty in 90 minutes. DNA evidence would later exclude him.
At the heart of McGee's case and others is whether Chicago police used their polygraph unit as a tool to obtain false confessions. At least five defendants — four of whom were charged with murder — have been cleared since 2002. In a sixth case, a federal appeals court threw out a murder conviction, leading to the release last month of a Chicago mother prosecuted in the death of her 4-year-old son.
A Tribune investigation found that Chicago police have long ignored voluntary standards for conducting polygraph exams, even as those methods and the examiners themselves have factored into cases costing the city millions of dollars in damages.
Court records, department documents and interviews show that Chicago police polygraph examiners have not followed key standards as published by national industry groups when administering the exams, which have long been controversial. The Chicago examiners' results don't have to be reviewed by a second examiner or supervisor. The unit has no continuing education requirements in place. And it records only a fraction of its polygraphs.
For decades, the department did not use numerical scoring for the tests, even though such scoring is strongly recommended by major industry groups. Police said in 2012 that they had recently moved to numerical scoring.
Also, one of the unit's three examiners said in a sworn deposition that he has not always taken notes in interviews before the tests, though state law requires it.
Mark Handler, the American Association of Police Polygraphists' research and information chairman, has not studied Chicago's polygraph unit but expressed concern that any department would disregard standards.
"By not following the standards, you place yourself at risk for errors, which can lead to an increased risk for a false confession," he said. "It's a very precipitous slope and a dangerous game to play because the ultimate harm is convicting an innocent person of a crime they didn't commit."
Chicago police officials said they follow state law and use validated techniques for doing polygraphs, though they would not say what techniques they use.
Cmdr. Joseph Murphy, who oversees forensic services, the division under which the polygraph unit falls, said polygraphs have long been an important tool, particularly when detectives hit a dead end or need to check the veracity of a witness or suspect. He stressed that examiners tell their subjects the exam is voluntary, advise them of their rights and get consent before proceeding.
"Our guidelines to the examiners are to provide the best service they can, be professional, be fair and impartial," said Murphy, who is not himself a polygraph examiner. "You want to get the best evidence you can and the right evidence."
Though polygraph results are rarely allowed in state court partly because of concerns about reliability, many experts believe the tests can be useful, providing investigators with information, for instance, that can then be corroborated. In Chicago, the unit's written policy states that although polygraphs should be used as an investigative aid, they are not a substitute for thorough investigation.
But in the cases in which Chicago murder suspects went on to be cleared — some after spending years locked up — police polygraph examiners were accused of making up a confession, using "trickery" to get an admission and telling a suspect he failed a polygraph that an outside expert would later deem too poorly administered to determine its result.
In five of the six cases, suspects were taken to Bartik. Suspects said they were drawn in by the promise of the polygraph — which they believed was a scientific test that would prove their protestations of innocence were true. Instead, they allege in court documents and interviews, Bartik obtained confessions from them by berating them, threatening them and lying to them.
Bartik said in court testimony that he got more than 100 confessions in a five-year period in pre-test interviews — a number that strikes some experts as extraordinary. They said the pre-test is not a time to try to get confessions; it's when the examiner explains how the polygraph works, gets consent and reviews the questions.
Craig Futterman, who represented one of the suspects who later sued the city, said he believes Chicago's polygraph unit has a systemic problem. "They're using the polygraph as an interrogation prop to try to overcome somebody's will and extract confessions, false or otherwise," said Futterman, a University of Chicago law professor.
Bartik did not respond to requests for comment.
In trials as well as in sworn depositions, Bartik has denied allegations of wrongdoing. Since he joined the police force in 1988, he has received numerous commendations, including more than 40 honorable mentions.
Murphy declined to comment on cases Bartik handled, but he called Bartik an excellent examiner — "probably one of the most knowledgeable polygraph examiners I've ever met in my over 40 years on the job."
The debate over whether polygraph exams are science or quackery has raged since one of the earliest devices was developed in California in the early 20th century and later honed in Chicago. Polygraphs remain a common part of criminal investigations in Chicago and in departments throughout the country, as well as federal law enforcement agencies.
Even if the results are rarely admissible in state court, the confessions obtained by an examiner or by detectives after a polygraph are.
Testifying as part of a lawsuit in 2010, Bartik said the 111 people who confessed to him during the five-year period ending in 2003 did so not after a polygraph, but before, in a pre-test interview.
Dan Sosnowski, who testified in the same lawsuit for the plaintiff, estimated in court that in his 30-year career that started in Chicago and took him across the globe on behalf of the U.S. government, no more than five suspects confessed to him during a criminal pre-test interview.
A past president of the American Polygraph Association, Sosnowski testified that Bartik's record of pre-test confessions was "unprecedented" and "highly improbable." He concluded that either Bartik used a "very accusatory, heavy-handed coercion method or (the confessions were) fabricated."
Illinois law prohibits a polygraph examiner from initiating an interrogation "for the purpose of eliciting a confession" during a pre-test interview. A polygraph examiner is supposed to remain objective and nonaccusatory. Only after the exam can an examiner interrogate a suspect.
Bartik has said in court records that he never interrogated a suspect during the pre-test interview, and his attorneys have argued that there is no evidence he fabricated any confessions.
Sgt. Chris Germann, who supervises the Los Angeles County Sheriff's Department polygraph unit, said that while confessions more commonly follow a polygraph, they can occur during the pre-test.
"If someone is good with people, and they're able to talk to people, yeah, people will confess to some people faster than others," he said.
Bartik has also said in court documents that there are times when he does not take notes during the pre-test interview, a red flag to Fred Hunter, a retired Hinsdale polygraph examiner who had served as a coordinator with the state agency that regulates examiners.
"This is basic polygraph procedure that is elementary to the law," Hunter said. "… If you don't have the document, not only are you in violation of the law, it's open to speculation that you didn't do what you were supposed to do. ... It seems to me (Chicago police) are kind of flying by the seat of their pants."
Murphy would not explain what specific technique his department uses for forming its questions or the order in which it asks them. Because even word choice or inflection can affect results, national standards dictate that examiners use techniques that research has shown provide accurate results.
In response to a Freedom of Information Act request, Chicago police said they had no existing public documents that named or described their polygraph techniques or analysis. The department denied requests for documents from polygraphs conducted in 2012. Those documents could show the unit has moved to numerical scoring of exams.
Murphy said his unit's polygraphs are peer-reviewed at least weekly, and examiners often review each other's work informally. He said that since 2005, when a new law required police to record a homicide suspect's entire interrogation — not just the confession — polygraphs in those cases were recorded and reviewed.
Of the 402 polygraphs in 2011, 85 were homicide cases, he said. The rest were not recorded.
"If you're going to do a polygraph test, one of the components should be video recording," said retired state Sen. John Millner, R-Carol Stream, a former Elmhurst police chief and longtime polygraph examiner. He said he began audio recording and numerically scoring his exams in the late 1970s.
Video recording might have made a difference in McGee's case.
After McGee's neighbor, Ethel Perstlen, had been found stabbed and burned in her bathtub in 2001, he agreed to take the polygraph, thinking, "'I'll do the lie-detector test and prove what I'm saying is true, and they'll let me go,'" he said.
Alone with Bartik, McGee said he asked about the exam. "He told me that we're not taking the test because he already knew how I committed this crime," said McGee, who claims Bartik interrogated him. His time in the polygraph room was not recorded
McGee spent 31/2 years in jail awaiting trial before he was cleared.
DNA from a dark brown stain on a telephone found off the hook later excluded McGee.
Bartik testified that McGee confessed, which is why he didn't administer the polygraph, and that McGee then repeated the confession to the two detectives on the case. City attorneys argued that the confession contained details the officers could not have known.
McGee sued Bartik, the two detectives and the city, alleging they fabricated his confession.
In 2010, a jury awarded McGee $1.3 million. As part of that total, Bartik and the detectives, in a rare move, were ordered to personally pay $110,000 each in punitive damages. But the city appealed and was recently granted a new civil trial because of a procedural error.
"Know that every person who comes to Bartik has already been interrogated by detectives with experience," said Russell Ainsworth, who represents McGee. "They interrogate and they use the best techniques they know, and the person has not confessed."
Despite being taken to the polygraph room, Lamar Blount also was never given an exam while under suspicion in a home invasion and murder in 2003.
Blount alleged in court records that Bartik tricked him into falsely confessing during the pre-test interview by telling him he had to admit being involved in the murder before he could take the polygraph exam. During the trial, a Cook County judge acquitted Blount without even requiring him to put on a defense.
Bartik is one of the unit's three full-time examiners. In two separate cases that led to a settlement, each of the other two examiners gave polygraphs.
Tina Figueroa-Mitchell was sued in 2006 for doing a polygraph inquiring about an infant's injuries that led to the Department of Children and Family Services wrongly removing the girl from her parents' custody for more than 8 months, according to a lawsuit.
She concluded that the parents failed the polygraph asking about their daughter's broken bones, but she did not numerically score the exam or conduct the mother's test in her native language, court records show. She also testified that she turned off the blood pressure cuff for part of one exam, though state law requires its use.
The city denied she acted improperly and argued that the polygraph was not an influential factor in DCFS' decision. The city settled with the parents for $2.5 million. Figueroa-Mitchell declined to comment for this story.
The third examiner in the unit, Kevin Howley, determined that Corethian Bell failed a polygraph asking if he killed his mother. Bell then falsely confessed and spent 17 months in jail until DNA evidence led to his release. He sued and received a $1 million settlement from the city, which did not admit wrongdoing. Howley did not respond to requests for comment.
Bell and at least one other suspect said being told by Chicago police that they had failed what they believed was a scientific exam weakened them into falsely confessing.
Bell, then 24, was already vulnerable to pressure. He is mildly mentally disabled, and his interrogation in 2000 stretched over 50 hours and, Bell alleged, included threats, accusations and physical abuse by the officers, according to court records.
Charles Honts, a Boise State University professor and licensed polygraph examiner, issued a blistering report on the case for Bell as part of his lawsuit. He wrote that "the primary motivation for conducting this examination was to use a failure on the polygraph as a tool to extract a confession" and that the examiner was "clearly in violation" of published standards on several critical points.
One error was Howley's failure to use structured numerical scoring, Honts said.
In a 2005 deposition in Bell's suit, Howley said he never actually scores a test using a formal scale.
"That's just the way we do it. We globally assess it," he said.
"… Purely you eyeballing it?" Bell's attorney asked.
"Yes," Howley answered.
The city agreed to Bell's $1 million settlement the next year.
Experts say it's hard to overstate the despair that a truthful person experiences in the face of a polygraph they are said to have failed. Research, meanwhile, has shown that the innocent are more likely to agree to a polygraph.
"What makes the polygraph particularly powerful is that it's couched in science," said Saul Kassin, a psychology professor at John Jay College of Criminal Justice in New York. "When they say you failed the polygraph, and you think it's infallible, it has a very strong effect."
Nicole Harris, whose conviction for her son's 2005 death was vacated by a federal appeals court in October, was released last month after more than seven years in prison.
She stepped into the polygraph room less than 24 hours after her son Jaquari was found with an elastic bedsheet cord wrapped around his neck. It was there, she said, that Bartik told her she failed the test — a charge Bartik later denied. The exam was inconclusive, police records show.
"I was shocked," Harris said in an interview. "... It didn't make any sense to me because I knew I wasn't lying."
As in other cases, Bartik's role in the confession is vehemently contested.
Harris testified that, in the polygraph room, Bartik called her a "monster" and told her if she didn't cooperate, she would spend the rest of her life behind bars.
"'You were angry, Nicole, you were angry,'" Harris testified Bartik told her. She said he accused her of wrapping the cord around Jaquari's neck.
"… I was just crying and crying, and I am just shaking my head," she continued from the witness stand. Then, she said, "I started agreeing."
Bartik denied Harris' account in court, saying he explained his role to her, had her sign a consent form, administered the polygraph and confronted her afterward with information that was contradictory to what she had told him.
She repeated the confession on video, saying she was so angered by Jaquari's incessant crying that she silenced him by wrapping the bedsheet cord around his neck. A jury found her guilty of first-degree murder in two hours. She was sentenced to 30 years in prison.
The 7th U.S. Circuit Court of Appeals would later call Harris' confession "by far the most damning evidence" — but one that the jury had "reasons to question," given that it followed 27 hours of interrogation, which Harris said included threats and a sleepless night, shortly after she lost her son.
In a unanimous decision, the justices said that had her other son, then 5, been allowed to take the stand, his testimony would have buttressed Harris' claim that Jaquari died accidentally while playing with the bedsheet cord.
The state has asked the U.S. Supreme Court to review the case. The state's attorney's office also could retry her.
Steven Drizin, legal director of Northwestern University's Center on Wrongful Convictions, which now represents her, contends that, among other police missteps, Bartik was central to getting Harris to confess.
"He was the straw that broke her back in many, many ways," Drizin said.Copyright © 2014, The Baltimore Sun