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.

They gotta make a living somehow!

500 bucks is a bargain if you can buy a "get out of jail free" card!

TC

Of course you should mention that Levy's Attorney refused a second polygraph in Guandiques Native Language within two months of his original polygraph after the results were called inconclusive due to language barrier problems. If the suspect test was inconclusive then so was the informant test for the same reasons

You might mention that Guandique told one informant Condit paid him $25,000.00 to kill Levy, told others that he was a spur of the moment attack assisted by two teenagers, others that he stalked her and still others that he didn't kill her at all. He told one of the governments "Reliable" informants that he left a knife in the body, saying that he intended to retrieve it but never got the chance. You say "hey wait a minute, they never found a knife!! No Problemo, he also told an informant he strangled her to death. He appears to be having difficulty resolving his M.O. (All of his confirmed attacks have been at knifepoint) with the phyical evidence (Broken Hyoid bone more indicative of strangulation). He kept a magazine picture of levy on his wall in prison. It is pretty clear for whatever reason wants his prison buddies to believe he committed  high profile murder, probably to enhance his "Yard Cred", but he certainly doesn't want any long term consequences.

You might also want to note U.S. Attorney Jeffrey Taylor said that there is no physical evidence linking Guandique to the crime.

Considering the lack of physical evidence, his ever-changing story, and the unresolved polygraph, there is a strong possibility that Guandique is a nutcase a'la Henry Lee Lucas and Ottis Toole who wanted people to believe they were the most prolific serial killers of the 20th century just for special privlidges and to keep their names in the paper. Back when Lucas was "Famous" he wouldn't talk to an investigator unless they brought him two cartons of cigarettes and a strawberry shake.

It will be interesting to see if Guandique is convicted. But I don't guess you really care about his guilt or innocence. You just want to criticize polygraph. I'm sure you will spin your story to make it appear that polygraph is somehow at fault either way. Won't you?

Its too bad that there isn't a Anti Lazy D.C. Cops website. If this case fails to be resolved it's their fault. They walked withing 10 yards of her body within days of her disappearance but allowed themselves to be deterred by rough terrain from doing a good job of searching.

wrote on Mar 4^th, 2009 at 3:06pm:

Of course you should mention that Levy's Attorney refused a second polygraph in Guandiques Native Language within two months of his original polygraph after the results were called inconclusive due to language barrier problems. If the suspect test was inconclusive then so was the informant test for the same reasons

Could you clarify what you mean by this? To begin with, rather than "Levy's attorney," I assume that you meant "Guandique's attorney." (Levy was the victim.) On what basis do you state that it was "language barrier problems" that caused Guandique's polygraph results to be deemed "inconclusive?" And why would that necessarily make the informant's results inconclusive?

I've seen no reporting explaining how it came to pass that that Guandique's polygraph results were changed from "passing" to "inconclusive." It seems that this was done only after Levy's body was found in the area where the informant who failed the polygraph said it was. I strongly suspect that this fact strongly influenced the later decision to characterize Guandique's polygraph results as "inconclusive."

I agree with you that the evidence against Guandique seems sketchy, and I'm not saying he's guilty. The point of my post was to call attention to the irregularity surrounding the reclassification of Guandique's polygraph results and the role that investigators' misplaced faith in the pseudoscience of polygraphy played in the investigation. Guandique was wrongly excluded as a suspect in large measure because he passed a lie detector test (while the informant who accused him failed one).

Those who would like to read more about the use of the polygraph in the investigation should see Chapter 10 of Washington Post staff writers Sari Horwitz, Scott Higham, and Sylvia Moreno's series, "Who Killed Chandra Levy?":

http://www.washingtonpost.com/wp-srv/metro/specials/chandra/ch10_1.html

OK Guandique's Attorney.   My Bad

As a language expert you should know that interrogatory concepts don't always translate verbatim between languages.

When  Guandique was asked Were you involved in Chandra Levy's Disappearance? (According to the link you postred?) A translator might have used any of the following words for "involved" even though each have a slightly different meaning.
implicado
relacion
involucrado
participar

Lack of a clear understanding between the polygrapher and the examinee as to what the questions actually mean can result in the wrong question being asked or in the examinee failing to understand the intent of the question.

Word meaning and understanding generally suffers when filtered though translator. I'm sure that you are aware that Hebrew scholars differ on translation of the Torah and I'll bet you have disagreed with other linguists who have translated some of the same documents you have worked on. I'll bet you have reviewed passages that were almost impossible to translate into english with the appropriate meaning.

If you use a polygrapher who speaks the native tongue of the examinee, you remove the filter and should be able to reach a clearer understanding because when you ask the examinee what the question or words mean to him you get an unfiltered response.
No matter how the charts might look, the results are inconclusive unless you can establish with some reasonable certainty that the meaning of the questions wasn't lost in the translation. None of the quotes, including the article you quoted, that I've read say that the charts scored in the inconclusive range. They have said the test was inconclusive. Under the circumstances I have described it would be possible for the charts to score as deceptive or non-deceptive, but the test would still be inconclusive due to problems with question content brought about by using a translator.  If someone asked the polygrapher how he knew the subject understood the questions and he was unable to provide a satisfactory response, the results of the test would be deemed inconclusive.

Presuming that the same polygrapher and translator were used for both the informant and the suspect then any defect in one examination would possibly be present in the other examination for the same reasons.

Using the same set of circumstances it wouldn't be proper to say one test was good and the other was bad unless you could eliminate the likelihood of misunderstanding from the test you are claiming is conclusive.


If you have  seen no reporting explaining how it came to pass that that Guandique's polygraph results were changed from "passing" to "inconclusive why would you conclude that it only occurred after the ody was found?  

I have read so much on this case over the past couple of weeks I can't recall where and can offier no link, but I do recall reading somewhere that Guandique was offered another test within 2 months of his February 4th initial polygraph. If this is true then the second test was offered before Levy's body was found. However I cannot offer this as fact.

If I can find that link I'll post it here.

Ed,

I understand well that nuances can be lost in translation. But I haven't seen any reports suggesting that there was any problem with the interpreter. On the contrary, then DC chief of police Charles H. Ramsey expressed no dissatisfaction with the interpreter. According to a Washington Times report dated Tuesday, 1 October 2002:


Quote:

The Washington Post reported in Sunday's editions that police were concerned that the polygraph test may be unreliable because the FBI administered it through an interpreter. Chief Ramsey dismissed that concern, saying he was satisfied with the use of the FBI-qualified interpreter. He said the same interpreter was used during the polygraphs of Guandique and the other inmate.

Perhaps there is reason to doubt the quality of the interpretation. But no specific problem has been alleged. I see no basis for assuming that it was an interpretation problem that resulted in Guandique's polygraph reports being deemed "inconclusive."

The reason I say that Guandique's polygraph results were deemed "inconclusive" only after Levy's skeletal remains were discovered in Rock Creek Park (where Guandique had assaulted female joggers) is that in late September and early October 2002 -- some five months after Levy's body had been found -- the Washington Post and Washington Times were reporting that Guandique had "passed" the polygraph.

So....

Inmate passes polygraph in murder case that brought down congressman

Inmate no longer considered suspect

Seven years later inmate is fingered

Media lies, reports polygraph administered seven years prior was inconclusive

For seven years the family of the victim has suffered

Polygraphs suck yet must be protected to sustain the lie

Ummm....FAIL

McClatchy DC reports that Ingmar Guandique is to receive a new trial in connection with the death of Chandra Levy:

http://www.mcclatchydc.com/2015/05/22/267597/as-key-hearing-on-chandra-levy.html

The Washington Post reports that prosecutors have dropped charges against Ingmar Guandique in connection with the murder of Chandra Levy:


Quote:

https://www.washingtonpost.com/local/public-safety/2016/07/29/7bccd898-55a9-11e6-bbf5-957ad17b4385_story.html Secret recordings emerged and the Chandra Levy case rapidly unraveled By Keith L. Alexander and Lynh Bui July 29 at 8:56 PM This week, prosecutors and defense attorneys were set to travel to a California meeting, preparing for the retrial of the man charged with killing Washington intern Chandra Levy in 2001. But that travel plan changed abruptly and the murder case crumbled. Lawyers on both sides had learned about recent secretly recorded conversations with the man who was to have been the key witness against Ingmar Guandique, Levy’s alleged killer. On Thursday, prosecutors dropped all charges against Guandique, telling defense attorneys that new information had left them unable to prove their case. New details about how the case unraveled emerged Friday from defense attorneys, who said the case was closed shortly after prosecutors received recordings from a Maryland woman of conversations she had with star witness and jailhouse informant Armando Morales. The woman told both sides that Morales said in the recordings that he lied when he testified that Guandique had told him about killing Levy, according to the defense attorneys from the District’s Public Defender’s Service. The defense team never played the recordings because they believed that they had been made illegally under Maryland law, said Laura Hankins, general counsel for the Public Defender’s Service. But after receiving the tapes, prosecutors dropped the charges, citing “new information the government received within the past week.” Signs that the prosecution’s case was beginning to unravel emerged July 21 during a routine hearing in D.C. Superior Court. At that hearing — one of dozens leading up to the retrial — one of the prosecutors told the judge about a witness who contacted them and told them that she had “information” about Morales. By the weekend, the woman, Babs Proller, had told prosecutors and defense attorneys about her secret recordings of Morales. Proller, a part-time actress, said she struck up an acquaintance with Morales, who had recently been released from prison, in early July when she was living in a Maryland hotel. “Ms. Proller indicated to us that Mr. Morales had told her that he had lied when he testified against Mr. Guandique,” said one of Guandique’s attorneys, Katerina Semyonova. Proller made the same assertion in interviews with The Washington Post. ...

George W. Maschke wrote on May 23^rd, 2015 at 7:47am:

McClatchy DC reports that Ingmar Guandique is to receive a new trial in connection with the death of Chandra Levy: http://www.mcclatchydc.com/2015/05/22/267597/as-key-hearing-on-chandra-levy.html

Ingmar -- a very funny name for someone from El Salvador!  ;D

wrote on Mar 5^th, 2009 at 4:44am:

So.... Inmate passes polygraph in murder case that brought down congressman Inmate no longer considered suspect Seven years later inmate is fingered Media lies, reports polygraph administered seven years prior was inconclusive For seven years the family of the victim has suffered Polygraphs suck yet must be protected to sustain the lie Ummm....FAIL

Hopefully, the decision in favor of a new trial won't revive the political career of Condit!   :P