The Lie Generator: Inside the Black Mirror World of Polygraph Job Screenings
wired.com
wired.com
Fast forward a bit, to where I am undergoing a polygraph examination for the NSA. The exam made me uncomfortable and nervous, but I thought everything was going well. Except for when my interviewer came back and told me I was showing sensitivity towards the hiding crimes question. WTF? And when they do this, they're just giving you enough rope to see if you hang yourself with it. But I had no idea why (or even if) I was showing sensitivity to this question.
They called me in for a 2nd polygraph, this time I didn't show sensitivity to hiding crimes, and I figured I was good to go.
No. I get called into a 3rd exam (each exam was separated by a couple months, mind you). This time the interviewer told me "You did better at the hiding crimes question than I thought you would" W.T.F.?!?! The interviewer then left the room and came back with a manilla folder, from which he procures a piece paper which he reads that I had a suspected larceny charge back at home. I honestly had no idea what he was talking about until I remembered the gas station incident. But after I try telling him about it, he tells me that he doesn't believe me and that he thinks I stole that gas. This leaves me extremely flustered and the rest of the polygraph was a train wreck.
3 strikes and I'm out, my conditional employment with them was terminated.
What irks me the most though, is that when I got back home I retrieved the larceny report from the court house, and in that document the whole story was laid out and my account of the situation was corroborated. So what the hell? Why throw me through such a ringer?
Fuck the polygraph.
And was he wrong?
You pumped gas without paying. Did you acknowledge it as a criminal act or as good fortune or not at all? Look at it from the interrogators perspective.
So they questioned them about this thing they knew about and he didn't reveal it, so they thought he was lying even though he honestly didn't remember it.
It doesn't work out well for people who know how polygraphs work in general or for people who do not reveal things they know about (intentionally or otherwise).
On one hand I understand how a discrepancy in the background check would raise concern, but on the other hand I feel that the police report easily shows no wrong doing on my part, and that this "blemish" could have been cleared up at the 1st polygraph interview, rather than stringing me along for 3.
But the people you were dealing with, as a rule, don't believe in happenstance, or coincidence. And they absolutely live by attention to detail. I don't know much about your situation, but based on what you've shared, you tripped a lot of alarms.
Well then congrats to the NSA; they successfully weeded out a spazzy dork
Okay, sure, there was no reason not to go inside and pay cash, and ask what's up with the pump, and maybe tell them to put a sign up, when it's all of maybe $20, maybe $40 at stake for a single tank. But I really want to know the background story from the perspective of the gas station. Like, how bad was the hit on that malfunction for them?
I really have to wonder, like was it a Mom & Pop station, or was it some huge multinational chain? And how long did the incident last? Did they lose an entire underground tank's worth of gas in less than a week, or even one day, with no transactions posted? I see "out of order" signs all the time. How did station employees not notice, and disable the pump in time? How many people got scooped up in the drag net?
They must have lost at least six figures worth of gas, and if it wasn't an independent station, I can only imagine that multiple station attendants lost their job for not disabling the pump. For a gas station to open a criminal investigation, to the tune of police reviewing possibly more than 24 hours of footage, running plates for every car that skipped out on a pump's error message, and making phone calls to track down individuals to dispense a warning under penalty of criminal charges, it must have been a real mess, and a total fiasco for the station owner.
Considering the insurance required for handling hazardous materials in a motor vehicle context, where a sleepy trucker could send an entire 18 wheeler plowing into the pumps, destroying an entire station, or a rusty leaking tank having the same effect, it's surprising that they wanted heads to roll over a malfunctioning credit terminal.
Not long, I left the gas station and received a call from the police within an hour or two, and I kind of remember them telling me it was only happening that morning.
I only remember that it was a Sinclair station
Forgetting to pay at the pump once in your life can hardly be called criminal behavior.
How is that a criminal act?
So, given that you paid-up for the gas when requested, you didn't intend to permanently deprive the owner of it which suggests that technically it wasn't larceny and could have been sloughed-off as a misunderstanding. Given you were dealing with an NSA interview, sounds like they were playing a psychological setup game to determine how calmly you would respond under pressure in the context of an intimidating polygraph test. Seems like they weren't too thrilled with the responses they elicited. Would you really want to work for an organization that plays those sort of mind games?
No, I don't anymore. But at the time I was excited for the opportunity
With each other, on a regular basis, for little more than shits, giggles and office-politics? Probably not.
Which is why it's more likely that this mind game isn't just some office antic, but instead means to mimic some class of situations that some kinds of three-letter-agency personnel could be required to overcome.
I would like to work for an organization that pays me so I can buy food and pay rent.
The actual results from the machine are pretty much just a red herring, but that doesn't necessarily mean the machines have no valid uses. If use of a polygraph has helped tricked even one criminal into (truthfully) confessing or scared away even one person with malicious intent, then they're useful.
I wouldn't want a polygraph to be used for evidence in court, but I would want them to be used if I were interviewing for FBI/CIA/NSA. (But I definitely wouldn't want them to be used for firefighter and parademic positions.) I know it's a bit paradoxical that they can be both useful and a sham, but I think that's how law enforcement views them, too. This article presents a lot of issues with how they're used, but I wouldn't expect them to be phased out for a very long time (unless someone makes a version that actually has more scientific validity).
Do you support their use too?
Trickery and deception when it comes to medicine and personal health are never fair game. I think there's definitely a difference between police telling a suspect "we know you did it" (when they don't actually know) and a doctor telling a patient "take this and you'll feel better" (when the drug actually does nothing).
Also, polygraphs are a little bit more functional than homeopathic medicine: a polygraph can not only induce nervousness, but can also notice it in some cases (even if the nervousness may not be concerned around a particular question and may not necessarily indicate deception and even if there are false positives). This helps create a feedback loop which makes it more powerful than a pure placebo (e.g. a "lie detector machine" you hook someone up to which isn't powered on at all and does literally nothing).
What about when it is known to work?
https://www.health.harvard.edu/mind-and-mood/putting-the-pla...
Why?
As well as being known to be a form of trickery it's also the case that it is easy to teach someone to manipulate the results at will.
All you achieve requiring it for those organisations is that there will be a) some who are rejected unjustly thanks to a false positive, and b) the people you really don't want in there have another bit of flim-flam to reinforce their apparent legitimacy.
As long as the examiners, recruiters, interviewers, and higher-ups fully understand polygraphs are a trick and nothing more, I still think they serve a useful purpose for three-letter agency hiring.
six years later and hes still working for us, we get an order from that same municipality to overhaul the intercoolers on nearly two dozen cop cars. I called up the pool manager and asked about the polygraph, and his response was they use outside contractors to get around the fact they have no certified mechanics.
Is it simply because its the "standard" now, and bureaucrats don't want to stick their neck out by getting rid of it? Is the fact that it is a machine that has been around for awhile, regardless of efficacy, give people that much comfort? Or are enough people really that misinformed?
Its a bit like marijuana legalization coverage - its rare to find arguments for maintaining the status quo as opposed to getting rid of it.
Lie detectors, Sea Wolf submarines, facial recognition software, etc etc etc. Many security technologies lie on the spectrum from "impractical" to out and out "doesn't work". But we just have to get comfortable with them, because they aren't going away.
The linked article discusses a large number of cases of racial bias. Obviously if you want to be biased in your hiring, but you want to hide it, a polygraph is very convenient way to manufacture some cover. (Alternatively of course, some of the police departments may have had no such intention, and merely been the victim of biased polygraph operators.)
As for the recent Supreme Court fiasco, incidents 35 years in the past rarely turn up much hard evidence. A polygraph is, again, a very useful way to manufacture something that looks like evidence. (And that's true regardless of the truth of Ford's claims. Just because polygraph results are fake doesn't mean they aren't accurate sometimes!)
In short:
> Who likes them?
The people who commission them, because they can get the results they want, and the people who operate them, because it's a pretty well paying job.
The lie behind the lie detector.
Not having the money to do so.
Prospective employers don’t prefer litigious candidates. (seemingly regardless of merit).
Under what circumstances would a case like this even come to pass? From what I gathered from the article, these administrative hearings where applicants may plead their cases are not court proceedings but administrative hearings conducted by law enforcement officials.
Relief would likely come only under another course of action, like a civil rights case brought against the hearing committee (as mentioned in the article).
Unless you are referring to an action brought against the hearing committee on the basis that the test was improper because the person administering the test was untrained. At that point, I would imagine the court would not want to prescribe the hiring practices of local law enforcement and would defer to them (unless those practices could be shown to be discriminatory against some protected class).
Also, I didn't get the sense that these tests are administered by untrained HR reps, but I may have missed something in the article.
Modern fMRI technologies can tell if people are fabricating stories. There's actual science behind them.
I think there was some actual science behind the polygraph too, but just having actual science on poorly informed/motivated participants isn't really enough for tech that will have long term adversaries.
I can't recall if 1/3 or 2/3 of myth busters staff could beat the fMRI once prepared and motivated.
Initial experiments on disinterested subjects given no information about past experiments may have legitimately been about the same for both fMRI and polygraph (at the respective times when participants could have had no information)..
My guess is, word gets around, and "inconsistencies" is just the excuse they need. I'm not saying it makes it right, because next it's going to mere coincidence that a black woman had "inconsistencies" when applying. But in this case, I might be willing to let it go.
> This time he had failed the New Haven polygraph for something cryptically called “inconsistencies.” “[But] I’m not hiding anything,” he said at the hearing. “I was being straight and honest and I’ve never been in trouble with the law. I’m not lying about anything.”
His argument seems to be that all of the polygraph tests were consistently wrong and that he didn't do any of those things. This is consistent with later comments by other people in the article:
> While undergoing a polygraph examination for a position at an FBI field office in New Haven in 2010, a black man was told that his recollection of using marijuana only a few times in high school was showing as deceptive, and that he should change his answer. Later, he wrote: “I was convinced that [the examiner] may have made an assumption, based on a stereotype about African Americans and drug use, and used that stereotype to profile me. I also realized that what [he] was asking of me would reflect negatively either way—if I didn’t change my answer I was being deceptive, and if I did change my answer I was lying on my application.”
Regardless, that's about as far as I got because my two personal experiences with polygraphs tells me they're about on the same level as dowsing rods. "Have you ever used marijuana?"
"No", he said, higher than a kite having smoked a bowl an hour before the test. It was asked both times, passed both times.
"He had first applied to the Connecticut State Police and was failed for deception about occasional marijuana use as a minor. He then tried again with a police department in New Britain, where a polygraph test showed him lying about his criminal and sexual history."
I think the article's point is that the polygraph _operator_ decided that (perhaps not consciously), not the machine itself.
You misunderstood the article. There's no evidence to suggest he lied ever.
I've literally seen someone who supposedly wrote firmware for the space shuttle fail fizzbuzz.
Remember those people in college who contributed nothing to the group project, but still somehow squeked by with a C? Those people now have jobs, and some have impressive resumes, particularly from .gov contractors where less than mediocrity reigns unfettered.
Have you been involved in the hiring process? Most of the people actively looking for work don't have a job for a reason. It's not 'insulting' to see if you can even vaguely do the job in an interview.
There's only so much vetting you can do before the 'can you vaguely code your way out of a wet paper bag' step.
Also: you didnt used to work in Boulder recently, did you?
No I have never worked in Boulder.
The candidates you're trying to avoid are really good at passing a phone screen since they have a lot of experience at it, constantly looking for jobs.
Polygraphs can produce errors that may be used as a wrongful justification for an employer to dismiss your application.
...on the other hand
A fizz buzz application that produces errors _should_ be used as a justification for a potential employer to dismiss you entirely.