The Chandra Levy murder investigation appears to be yet another example of a case where misplaced reliance on the pseudoscience of polygraphy led to investigatorial misdirection.
Washington Post staff writers Sari Horwitz and Scott Higham
report that on 3 March 2009, a warrant was issued for the arrest of Ingmar Guandique, who was polygraphed about the disappearance of Chandra Levy in 2001. Regarding that polygraph examination, Horwitz and Higham write:
Quote:Guandique, a day laborer, came under the scrutiny of investigators months after Levy disappeared, but a series of delays and missteps allowed the case to languish. Nine months before Levy's remains were found in Rock Creek Park, a D.C. inmate came forward to say that Guandique had confessed to the crime while they were in jail, but the inmate's account was dismissed after he failed an FBI-administered polygraph exam. A polygraph test taken by Guandique before he was sentenced in the other two attacks was deemed "inconclusive." Neither exam was administered by a bilingual polygrapher, even though Guandique and the other inmate speak little or no English. Polygraph results can be skewed if there are translation problems, experts say.
Interestingly, while it is now being claimed that Guandique's polygraph results were "inconclusive,"
this is not what was reported back in 2002. Press accounts from the time make it unambiguously clear that Guandique
passed a polygraph test, and that investigators relied on those results in ruling out Guandique as a suspect. (At the time of Guandique's polygraph test, then Congressman Gary Condit, who had had an extramarital affair with Levy, was the focus of the investigation, and polygraph results were also at issue. See
FBI Doubletalk on Condit's Polygraph Results.) Here is what
Washington Post staff writers Sari Horwitz and Allan Lengel
reported on 29 September 2002 regarding the polygraph examination of Ingmar Guandique:
Quote:Detectives in the Chandra Levy murder case are focusing on a man convicted of assaulting two women jogging in Rock Creek Park last year -- a suspect who was initially discounted after he passed a polygraph test that investigators now believe was flawed.
Ingmar A. Guandique, 21, has been in prison for the assaults on the joggers since July 2001, two months after Levy disappeared. After her remains were found in the park May 22, some investigators reexamining his case were struck by the similarities in the three crime scenes, law enforcement sources said.
Investigators then discovered that a Spanish-speaking interpreter instead of a bilingual polygraph technician was used in administering Guandique's polygraph, sources said. Relying on an interpreter, according to legal experts, can skew the results of the test because the questions are filtered through and possibly altered by the interpreter.
...
A Tip Discounted
D.C. police first spoke to Guandique about the Levy case in the summer of 2001 after U.S. Park Police alerted them to his arrest in the jogger assaults, according to court records. But law enforcement sources said they found nothing to indicate he was involved in her disappearance, especially since, at the time, they weren't aware that her body was in the park.
After Guandique's arrest, an inmate at the D.C. jail told authorities that Guandique had confided in him that he stabbed Levy and left her body in the park, law enforcement sources said. The inmate didn't try to trade the information for a lighter sentence, saying he came forward because he felt bad for the Levy family.
In September 2001, the inmate failed a polygraph test, also administered through an interpreter. Guandique, who denied involvement in the Levy case, passed, the sources said, and authorities felt comfortable that he was not their man.
When Levy's body was found eight months later, Guandique's name surfaced as someone who had attacked other women in the park. High-ranking police, knowing that their detectives had discounted him because of the polygraph, played him down as a suspect, with Ramsey scolding, "The press is making too big a deal of it."
Ramsey's then-deputy, Terrance R. Gainer, was more blunt: "He wasn't our suspect then. He's not our suspect now."
Ramsey last week defended the use of the interpreter. "When you've got language issues, it's not unusual to use a translator," he said.
But Billy Franklin, director of the Virginia School of Polygraph in Norfolk, said he prefers not to use interpreters because if they don't pose the questions correctly, the answers can be wrong.
"In such an important case, they should have used a bilingual examiner if possible," he said.
James Starrs, professor of law and forensic science at George Washington University, contends that because lie detector tests can be unreliable, they shouldn't always determine the course of an investigation.
"Simply because someone passes the test, they shouldn't be written off, absolutely not," he said.
The re-characterization of Guandique's polygraph results as "inconclusive" and the suggestion by "experts" that "polygraph results can be skewed if there are translation problems" seem to be post hoc rationalizations for an apparent failure of the polygraph. But there is a simpler explanation for such failure:
polygraphy is junk science. No one should be surprised when an invalid test produces erroneous results.
Pages: 1
The Polygraph Examination of Chandra Levy Slaying Suspect Ingmar Guandique (Read 27027 times)
