By Erin Sherbert
By Erin Sherbert
By Leif Haven
By Erin Sherbert
By Chris Roberts
By Kate Conger
By Brian Rinker
By Rachel Swan
Editor's Note: In the opening paragraph of this column, Matt Smith listed theoretical examples of "uncomfortable questions" that could elicit a physical response. Smith did not actually pose those questions to polygraph examiner Patrick Coffey, nor did Coffey ask his test subjects any such questions.SF Weekly regrets any confusion.
The way polygraph examiners such as Patrick Coffey see it, when a person is pressed with truly uncomfortable questions — Are you racist? A dissembler? A bully? — beads of sweat will appear on his palms. His breathing will quicken, or just as suspiciously, slow down. His heart will beat faster. His blood pressure will rise, and the sensors strapped to this deceptive person's body will feed results to a computer armed with an algorithm that will show he's a liar.
According to popular advice on cheating the lie detector, the key to passing is to stay calm. No matter how uncomfortable you are, breathe normally. Pretend you're in an ordinary school exam; remember mind over matter.
Coffey doesn't believe the polygraph can be easily beat. It requires remaining preternaturally calm while being asked tough questions. And that's how I would describe his demeanor when I recently cross-examined him about his views on Muslims. When I asked him whether he was a bigot, I could feel my own throat tighten with nervousness. Coffey's radio baritone, however, continued with the same even patter he might have used to recite a recipe.
"Don't try to back me into that kind of corner," he said. "I grew up with an Asian mother. I learned the Vietnamese language from a very early age."
I came away from listening to Coffey's mellifluous voice thinking he may have been telling what he considered to be the truth. Few people, after all, believe themselves to be prejudiced. But I wondered whether this professional lie detector might be fooling himself.
Last week I wrote about how the San Francisco Police Department still uses a lie detector to screen recruits, despite the fact that scientists have roundly debunked its effectiveness. Using such a discredited method to help choose cops, I noted, could disqualify the best prospects and give a passing grade to skillful liars.
I also noted that Coffey, whom the city paid $81,463 during the last fiscal year to conduct the screening tests from his Burlingame consultancy, believes scientific criticism of his favored method is unfair.
But what I didn't describe is what a peculiar choice the SFPD has made in selecting Coffey as the one who would conduct what scientists agree is a hit-or-miss method for determining whether aspiring cops are truthful when they say they should be trusted with guns.
In the spring of 2005, on www.antipolygraph.org, a Web site dedicated to scientific debunking of the polygraph, Coffey wrote of prominent polygraph critic George Maschke, a Ph.D. linguist who works as a Farsi translator in the Hague, "I doubt even without the polygraph that you could now meet security criteria to serve in any capacity given your choice to 'work' in socialist Holland, which like France is losing it's idenity [sic] to Islamic Immigration there." In a later post on Maschke's site, Coffey wrote, "George should stay in Holland or some other Socialist nation. He is apparently more comfortable in a nation like that, or France, which has lost its/their respective identities to massive Arab/Islamic immigration."
Maschke said that during 2005 Coffey lurked on his Web site for some time under various aliases, making caustic comments. "He basically became quite a troll on our message board, and I eventually banned him, and he came back under a variety of monikers," Maschke said, adding, "He's got a lot of biases coming to the table. To have someone with that mentality screening police applicants is inappropriate."
I asked Coffey about his statements. He said he was merely making a casual observation about changing times: "The Holland or France that you or I might have visited in our childhood might not be the Holland we would have visited years ago," so Maschke "may have a different perspective than the average man on the street here, because of his interaction with these people on a personal and professional level. Compared with the man on the street here in the U.S., what he feels is reality might not be that of the average person here."
I have no specific evidence that Coffey's feelings about Muslims, Arabs, or even Dutch people have tainted the way he screens applicants for the SFPD. But the issue should be raised because the statistical accuracy of a polygraph exam, according to a 2002 National Academy of Sciences study, is only slightly better than random guesses in identifying potential liars or truth-tellers. This large accuracy gap leaves plenty of room for bias.
"The potential for discrimination is certainly there," said David Faigman, a UC Hastings law professor who served on the National Academy of Sciences polygraph panel. "What we do know is there's a high degree of subjectivity and a high degree of human interaction relevant to the test. When you have those two things, you have to be concerned that bias or prejudice manifests itself."