Can I beat a lie detector?

Humiliating myself for fun and profit.
Jan. 27 2005 6:25 PM

Can I Beat a Lie Detector?

I always tell the truth. But not today.

Illustration by Nina Frenkel

I am sitting on a pad that sends an electronic signal to a computer every time I move my body. Circling my chest and abdomen are two rubber tubes, the pneumographs, which monitor my breathing. Tiny metal electrodermal plates secured to the index and ring fingers of my right hand with Velcro strips are measuring how much I sweat. A blood-pressure cuff is inflated tight around my left arm. Interrogating me about my moral failings is Darryl DeBow, owner of the Virginia School of Polygraph. I am lying my head off to him, just like—depending on your politics—Bill Clinton or George W. Bush.

Emily Yoffe Emily Yoffe

Emily Yoffe is a regular Slate contributor. She writes the Dear Prudence column. 

I'm also flexing my anus as if I'm in a proctological triathlon—but more on that later.

For this "Human Guinea Pig," the column in which I do stuff you'd rather not do yourself, my plan is to see if I can beat a lie detector test. Although I'm a person of impeccable ethics, the whole idea of taking a polygraph exam, let alone trying to outfox it, has me in a state of fretfulness resembling Lady Macbeth in the fifth act. (OK, maybe I more resemble Ben Stiller facing Robert De Niro in the polygraph scene in Meet the Parents.)

To prepare myself, I read online advice on how to fool the machine. I consulted the comprehensive Antipolygraph.org, co-founded by George W. Maschke, who became an activist when he was turned down for a job with the FBI because of what he says was a false finding of deception during his required polygraph. I also spent $19.99 for a seven-page manual from Polygraph-test.net, which promised, "With the techniques in this manual you will be able to fool the machine so that you can successfully pass your exam no matter what!"

First, I learned, I had to understand the types of questions I was being asked. Polygraphers generally ask both "control" and "relevant" questions. Control questions are an unnerving tour of your past transgressions. An example might be (not that I was asked this), "Have you ever noticed that the dog has taken a dump in the house but pretended you didn't so your spouse will clean it up?" The relevant questions are pointedly specific about what the exam is really trying to uncover. For instance (and not that I was asked this, either), "Do you know where Jimmy Hoffa is buried?"

The purpose of the control questions is not actually to find out whether you stick your spouse with the dirty work, but how your body responds to anxiety-provoking questions. It's the relevant questions the examination is really designed to answer. So, if you didn't have anything to do with Jimmy Hoffa's disappearance, your confident denials about his whereabouts will elicit a milder response than to the control questions. But if you were involved, and said you weren't, then your guilty knowledge should create Mount Everests on the graph.

My online research said that beating the machine was easy. During the control questions I simply had to send the needles into a frenzy. Then, when I lied on the relevant questions, the needles (actually the simulated needles—polygraphs have gone digital and the examiner is looking at a computer screen) would remain calm. All that was required to cause the frenzy was the activation of a powerful weapon already in my possession: my sphincter. As the $19.99 manual explained, "[S]lowly flex your anus." Antipolygraph.org somewhat more romantically described the action as an "anal pucker."

Both warned not to overdo it, or get my buttocks muscles involved, lest the gauge I was sitting on reveal my squeezes. If the examiner picked up that I was using this or any other countermeasure, the session would be ended and I would be branded as deceptive. To perfect my technique I needed to practice daily, occasionally sitting on my hands to make sure my buttocks remained inert. Though I did my exercises, I doubted that double agent Aldrich Ames beat all those CIA lie detector tests because he had a sphincter of steel.

Lie detectors have been controversial since they were first used during the early 20th century. Even the matter of who should be given recognition for the creation of the modern lie detector is a matter of dispute. According to the National Research Council, credit—or blame—belongs to psychologist William Marston, who made another contribution to popular culture that has also sent hearts racing for decades. He is the creator of Wonder Woman.

Marston was an expert witness in a 1923 murder case in which he argued that polygraphs should be admissible. An appellate court was not persuaded as to their scientific reliability, and ever since polygraphs largely have been excluded from trials. Despite this, the federal government and local law-enforcement agencies are ever-larger consumers of polygraph exams. They are used not only to try to make people confess to crimes, but also as an employee-screening tool.

To determine whether polygraph exams have any validity, the National Research Council conducted a major study that was released in 2002. The 398-page report is easy to summarize: Polygraphs are baloney. The report found that lie detector exams are so subjective and undependable—are they really measuring deception, or just fear, for example—that they are inherently untrustworthy.

Illustration by Nina Frenkel

Armed with this knowledge, and a pucker tighter than a baby sucking on a lemon, I go to meet DeBow. With his ramrod bearing and buzz-cut hair, he looks just like the former deputy sheriff that he is. I'm already intimidated. Since he's not trying to divine whether I kidnapped a Teamsters boss, he comes up with a clever scenario to test my honesty. He sends me upstairs where, he says, a man is sitting at a desk. I then have a choice of demanding the man give me money, or politely leaving. If I take the money, DeBow tells me, I will be committing a robbery. I walk up the stairs, take a $20 bill, hide it in my pocket, and return to DeBow.

Then we go into a little room with a computer monitor and the rest of DeBow's equipment. DeBow places me in a chair opposite him to conduct the standard pre-polygraph interview. I have read that since everyone has done something naughty, when I'm asked I should just admit my minor transgressions. I'm impressed that within minutes DeBow has moved from innocuous biographical questions (name, place of birth) to an exploration of my tortured family relationships and my thieving past. When he asks if I've ever stolen anything, I acknowledge that on about half a dozen occasions in my late teens I shoplifted magazines. He reacts as if I'm Ma Barker reincarnated. I try to laugh it off, but I feel guilty.

Then he asks me if I took the money from the man upstairs. I look calmly and directly at him and answer, "No." For about 45 minutes he reviews my childhood and my criminal inclinations, every so often interrupting these reveries to ask again if I stole the money. I keep feeling I should try to take charge of the interview, like Sharon Stone in her provocative, leg-crossing (and uncrossing) interrogation scene in Basic Instinct. Of course, to carry that off it helps to look like Sharon Stone.

Finally DeBow hooks me up. When he's done, I feel trussed up like a Thanksgiving turkey. He goes over the series of 10 questions he's going to ask me. It is obvious which ones are controls ("Have you ever committed a serious, undetected crime?" "Since being a journalist, have you ever lied to make a story more than it was?") and which ones are relevant ("Today, did you take that money from the man upstairs?" "Did you steal that money?").

As he asks each control question, I do my flexing and two other techniques the Web sites recommend: I slightly alter my breathing and perform mathematical sums in my head but answer honestly. In response to every relevant question, I lie and say I didn't take the cash. We go over the series of questions several times and by the end, even though there is no longer a Soviet Union, and even though I have never spied for it, I feel like confessing that I have.

DeBow has yet another series of questions he wants to ask me, but first he says, in an exasperated tone, that I need to stop doing my erratic breathing. I don't know if he doesn't mention the sphincter thing because I am such a master at it, or because he's too polite. I drop the breathing but keep the squeezing. His new questions are pointed. "If you stole the money was it a $100 bill?" "Was it a $10 bill," etc., through the denominations. I reply "no" to each. When he finishes this he turns to me and says, "You want to give me the money back?" I try to pretend I don't have it, but he assures me that I failed the exam. When I ask him to prove it, he shows me the response to when he asked if I'd stolen a $20 bill. My sweat glands were releasing a Nile of guilt.

DeBow agrees with the computer-generated score on my overall exam that there was a 99 percent chance I was being deceptive when I denied the robbery. But he says he hadn't even needed to hook me up to know that. During our pre-interview, he says, it was obvious when I was lying. During most of the pre-interview I was as twitchy as a picnicker sitting on a fire-ant nest. But I'd become uncharacteristically wooden each time he asked about the money.

He also says that if I were not a journalist he would have ended the examination within a few minutes because of all my stunts. He had noticed the sphincter thing, he says. He shows me it was hard to miss since every time I did it I created a spike on the computer resembling the Empire State building.

As we review the charts it turns out that when he'd asked me if I'd ever lied to make a story better and I said "no," my breathing, sweat-gland, and blood-pressure responses made me look like a combination of Jayson Blair and Stephen Glass. DeBow reassures me that just meant he'd come up with a good control question—one that got my anxieties flowing. But there was a bigger reaction when I actually lied. "You try to be a good liar," he says. "But you suck at it." I take that as a compliment.

Is there something you've always wanted to do but were too scared or embarrassed to try? Ask the Human Guinea Pig to do it for you. E-mail me your ideas at humanguinea@hotmail.com.