Washington Post staff writers Allan Lengel and Petula Dvorak report. Excerpt:
Rep. Gary A. Condit’s attorney announced yesterday that the congressman had passed a privately administered lie detector test that asked critical questions about the disappearance of Chandra Levy, but police officials immediately questioned its validity.
The announcement that Condit had passed a polygraph — administered by an expert hired by his legal team — caught D.C. police by surprise and publicly revealed the increasing tension between the congressman’s camp and the investigators who have spent 2 1/2 months on the high-profile inquiry.
D.C. police questioned the usefulness of the polygraph and suggested that Condit’s attorney, Abbe D. Lowell, had not played fair in the week-long talks over whether the Democrat from California would submit to an FBI-administered test.
“My impression was that we were going to continue that dialogue. I took him at his word,” Executive Assistant Police Chief Terrance W. Gainer said of the negotiations with Lowell. “I just didn’t expect it quite this way.”
Lowell, accompanied by a chart that detailed what he called Condit’s cooperation with investigators, said at an afternoon news conference that the examiner asked Condit the questions that “count.” The three were: “Did the congressman have anything at all to do with the disappearance of Ms. Levy? Did he harm her or cause anyone else to harm her in any way? Does he know where she can be located?”
Lowell did not list any other questions. He said the polygraph expert, former FBI agent Barry D. Colvert, concluded that “the congressman was not deceptive in any way and in fact had a probability of deception of less than one-hundredth of 1 percent to the only questions that mattered.”
Former FBI agent Barry D. Colvert’s conclusion that “the probability of deception was one-hundredth of 1 percent to the only questions that mattered” is without foundation. Polygraphy lacks both standardization and control. While it has an inherent bias against truth-tellers, outcomes can be manipulated by either the polygrapher or the subject. With such weaknesses, polygraphy can have no diagnosticity, and Colvert’s claim of certainty within one-hundredth of 1 percent is self-delusional nonsense.