Hi, I am Elle, and this is Quid Pro Quo, your friendly neighborhood forensics-like podcast. It's too late to change events, it's time to face the consequence. So strap on your lasso of truth and crank up the Depeche Mode, because today we're looking at the policy of truth, and by that, I mean polygraph testing. The Quid Pro Quo podcast deals with crime, psychology, and related themes.
Listener discretion is advised. I belong to a group on Facebook for a podcast called Podcasts We Listen To, and this has been a great resource for finding podcasts to listen to. On March 26th, Facebook user Jane Britt made a post in the group. I'm quoting the post here with her permission.
Jane said, true crime podcasts. I wish the scientific evidence which refused the reliability of polygraphs could be acknowledged alongside any polygraph evidence. I've heard polygraph evidence presented many times now, and only one podcast has discussed one part of the evidence currently available against its use. Thanks!
As a true crime junkie with a loud inner Jiminy Cricket, this is an issue which always frustrates me. I also have a bit of tangential personal experience with polygraph testing, so today's episode was a no-brainer. Long ago, in a bachelor's degree far far away, I signed up to be a research assistant on one of the psych studies on campus. The vast majority of what I did at first was phone calls to recruit subjects and audio transcription of the experiment recordings.
They didn't have a consistent system, and I'm a very fuzzy person, so I sort of took over the transcription side of the project. I worked on the study for several semesters for credit, and it was a good fit, so when one of their paid assistants graduated, they offered me a job. That's when I got to start working on the other side of the interviews I had been transcribing. This study brought in couples and recorded them while they had conversations about topics known to cause friction in their relationship.
They were also interviewed about neutral subjects to provide a baseline. They were video and audio recorded, but most importantly to this episode, their stress reactions were also recorded using breath monitors, skin conductance monitors, heart rate monitors, and audio visualization equipment. Ultimately, all of this data was compared to the actual content of what these people were saying, to study a number of things about how people in contentious discussions escalate and de-escalate emotionally, and what that looks like physically. Consistency is super important for psychological research, so it was pretty common for the people who actually performed the testing to grab an assistant to go through the motions of all the testing protocols.
So even though I've never been tested for anything, I have been hooked up to the full set of machines which are commonly referred to collectively as a polygraph. People think of a polygraph as a discrete machine, but in reality it is a collection of a lot of sensors. This was a while ago, but from what I remembered there was a camera in the room, a lapel microphone for audio recordings, a chestband to measure breath rate, a blood pressure cuff, multiple sensors on my hands to measure galvanic response, and one on my ankle which honestly is still a mystery to me. Maybe my pulse, I don't know.
If you're not familiar with galvanic response, the basic idea is that when you get nervous or excited, you sweat, and that extra salty liquid makes your skin conduct electricity better. A tiny, imperceptible current arcs across a portion of your body, usually the hand, and a sensor measuring changes in how skin conducts electricity indicates how much the person is sweating, which supposedly tells you how worked up they are. When we analyzed data from the interviews, we would see all of the spilogical inputs at once plotted over time and expressed as waves, much like the monitors you see hooked up to someone at a hospital. We would record any notable spikes or drops in these waves, and in the long run, this info would be compared to what they said in the interviews to see how what they were expressing physically matched up to the context of what they were saying.
Their facial expressions in the videos were also studied, but I was never personally involved in that part, so I can't speak to it. What we were measuring in the study paralleled lie detection in that we looked at what the person said and what their body said at the same time, and plotted out how changes in their discussion reflected changes in their bodies. Where it differed was that in our study, the context of the people's conversations didn't really matter beyond picking a subject that you felt made them fighty. We were not using the body data to make judgements about the content of the discussions.
What we measured basically worked out to was this person worked up at this particular point in time. However, in an actual polygraph test, the content of what the people are saying or reacting to is pretty important. One of the earliest versions of the polygraph was invented by William Molten-Marston, Harvard educated psychologist, lawyer and inventor, who comic book aficionados may recognize better, as none other than Charles Molten, the creator of Golden Age comic character Wonder Woman. Clearly, the men had a bit of a fixation with the truth.
Polygraph testing is based on the theory that people's bodies will unconsciously betray their emotions, that people will have a greater automatic physical response to something that makes them uncomfortable than to something which doesn't worry them. The problem is that there is no specific physiological response that reliably says, this is a lie, or this is the truth. All we can measure is discomfort. The theoretical framework underlying polygraphy requires the assumption that discomfort is indicative of guilt.
There is a really inexplicable logical leap here. The polygraph tries to gain validity by measuring a number of unconscious reactions at once. Poly stands for many and graph for writing, so a polygraph writes out several different results. Somehow all of these unreliable measures of discomfort when taken together are supposed to add up to guilt vultron.
Polygraph testing is more heavily used in employment circumstances than legal due to the difference in the standard of proof rules and level of potential personal ruin as a result. I am going to focus here on polygraph testing as it relates to crime, because that's how we roll here at Quidbroquo. There are probably as many different ways to go about a polygraph test as there are polygraphers, but there are a few commonalities. In all testing, a set of completely neutral questions with unambiguous answers like, where are you right now?
What color is this table? Are you seated? Provide the person's physical baselines at the time, and on a really mundane level, show the equipment is working. While there are a number of known protocols, the meat and potatoes of the questioning is generally approached one of two ways.
The first is a technique called the control question test. According to the American Psychological Association, control questions concern mis-deeds that are similar to those being investigated, but refer to the subject's past and are usually broad in scope. For example, have you ever portrayed anyone who trusted you? The idea is that if a person is innocent, they will freak out more at the broad questions than the specific questions related to the crime.
Presumably, everyone has had mis-deeds in their past, so a question about past mis-deeds is going to elicit at least a bit of a response. In contrast, if they have nothing to hide about the crime they are accused of now, supposedly they have no reason to react to questions about that crime. So the reaction to the past, nebulous mis-deed should theoretically be greater than that of the current crime in question. If the responses to the two types of questions are even, the results are considered inconclusive, and if they react more to the crime specific questions than the past control questions, the response are considered indicative of guilt.
As you might imagine, this completely ignores the possibility that the person being tested is innocent of a deed, but scared of something to do with the crime at hand. This may be a number of things, such as fear of the actual perpetrator or fear of false conviction. Likewise, it assumes that a guilty person will show fear or discomfort at the idea of being caught out, but that presumes they feel guilty or fearful in the first place, which is a pretty unfounded assumption. There are multiple techniques shown to effectively skew the fear responses in the tests, including certain forms of emotional coaching and even some mood-altering drugs.
In more extreme examples, as I discussed in the Fool Me One's episode, one of the significant features of anti-social personality disorder is a lack of emotional response to what most people would consider to be uncomfortable circumstances. Altogether, this means that the test may be especially prone to falsely flag nervous, innocent people and falsely clear, laid back guilty people, which is of course the opposite of what we want. An alternate approach to polygraph testing is what is called a guilty knowledge test. In this case, testers use multiple choice questions about the crime, where only one answer reflects the facts.
If that answer gets the strongest response, it's considered a sign that the person is hiding their knowledge about the crime. While this test is considered somewhat more theoretically sound, it requires that both the test-giver and the subject know information about the crime, which is not common knowledge. Having enough specific withheld information to make a solid attempt at this test is pretty rare and can be messed up by a lot of common circumstances. A positive result may reflect more about the investigator's poor question-framing or evidence handling than the subject's guilt.
One example may be an interrogator that led a detail slip during an interview. A negative result may also reflect not necessarily innocence, but instead ignorance, even in a guilty subject. As often investigators notice things about the crime, which the perpetrator did not, consider that if the perpetrators noticed everything that happened during a crime, investigators would have a very hard time finding evidence at all. A major limitation of the use of polygraph testing in criminal proceedings is research.
There are several legal frameworks for admissibility of evidence across the US, but for the most part they boil down to. Does this have solid enough scientific backing that we can base life-altering beyond reasonable doubt decisions on it? Polygraph research is so contentious that increasingly the answer in courts is, uhh, what else you got? Which is not what I'd call a shining endorsement.
I'm not going to cite statistics here because frankly, I don't think I need to. The research is all over the map, so instead I'll discuss the inherent problems in testing and hopefully that will make it clear why it's not even worth getting into the numbers. Remember how in my last episode I discussed how huge swaths of research are based on studies done on college students and are not very functionally generalizable to real life? I'm not sure you could find a more perfect example.
Keep in mind that this set of tests measures involuntary emotional responses. How exactly do you faithfully reproduce the emotional impact of dire, personal consequences in a controlled and repeatable research situation? How do you fake the responses of a person who is deep within the bowels of a police station lying to safer skin from a murder conviction they totally earned? How about a job applicant in a high-pressure career, make or break interview?
Or a completely innocent person facing police who don't believe them? Are you going to reproduce that testing on someone in no danger whatsoever in a safe, sterile environment? Even if you tested convicted criminals, do you think their responses in a consequence-free environment are going to be comparable to someone facing jail time or the loss of a job? Consider that the sort of criminals who are willing to be a part of a study may differ significantly from other portions of the criminal population in unclear ways.
Do the guidelines you develop for murderers results apply to petty theft or to reluctant witnesses, to drug users who are claiming to be sober, to pedophiles? Are results based on American offenders applicable to people brought up in other cultures? And to that, judging the effectiveness of people who administer and interpret the tests themselves, there is no one scientifically proven protocol and requirements for training vary from state to state. This is a research design nightmare made flesh.
According to the American Polygraph Association, scientific evidence supports the validity of polygraph examinations that are conducted and interpreted in compliance with documented and validated procedure. However, if you read their own 2011 meta-analysis which is available via an article in shown films, they can't even define what that means. It's spelled out in black and white in their accompanying discussion text that defining compliance and documented and validated procedure once and for all are goals, not past achievements. These extremely recent concerns were born out of dire economic necessity as polygraph testing falls out of favor in more and more applications.
When you read between the lines, this study is not undertaken to protect people undergoing polygraph testing. It is meant to protect the industry of polygraphy. There are considerable problems to overcome with the backing science and training procedures before the practice can be declared consistent and reliable enough to even form a centralized standard. Much less use that standard to create the body of research necessary to make a firm statement on its validity in the field.
The admissibility of polygraph testing and legal proceedings varies wildly by state. Here in New York, they are not allowed in any law enforcement or employment context for any circumstance, but many states have provision for their use in the investigation process, in parole hearings, or in courts if specific parties agree to its admission. I will provide a link with the detailed breakdown of these state requirements in show notes. According to the US Department of Justice page on polygraphs, linked in notes, neither the United States Code nor the Federal Rules of Evidence have a specific provision concerning the admissibility of polygraph examination results.
However, it seems to be heavily frowned upon, as there is such a high likelihood it will be challenged and eventually thrown out. When the Department of Justice basically says, polygraph results are probably not going to make it to trial, so don't even bother wasting our time with this. That's a strong statement of lack of faith. The OG science aside, I doubt the use of polygraph is going to fall entirely out of favor for a long while, as it does have other applications.
Although no state can force you to take a lie detector test, refusal can be viewed as evidence of guilt, much like we often see when someone makes a decision to lawyer up rather than talking directly to the police. There is some reason to believe that, regardless of the validity of the test itself, the threat of undergoing polygraph testing can be an effective tool to psych out a guilty party. Of course, researching this effect has potentially even more complications than researching the test itself, which is almost impressive from a theory standpoint. Officers will sometimes use the threat of exposure to take stock of a suspicious person or leave them to believe that the police have more on them than previously thought.
Per Sorrels VUS. Criminal activity is such that stealth and strategy are necessary weapons in the arsenal of the police officer. I personally heard one first-hand account from someone involved in the criminal justice system during the 70s of an interrogator who got a confession by telling the suspect that they were undergoing a polygraph test. Instead of the usual arrangement, they put the suspect in the room with a disguised photocopy machine set up beforehand to print the word liar.
They phrased their questions well, and every time they pressed the copy button, the lie detector made its judgment sneaky for sure, but perfectly legal. There is tentative evidence that a person who has already admitted guilt may be more likely to give expanded detail in the context of a polygraph examination, as the framework of the examination may let them feel like they are off the hook for making the decision to tell the truth. But this has nothing to do with the test's ability to extract truth and everything to do with the subject's attitude about unburdening themselves. If you are interested in an interview technique which does show promise at detecting falsehood, but in a very different way than polygraphy, I highly recommend episode 2 of the podcast Criminal titled Pants on Fire.
I will provide a direct link to this episode in show notes. If you want to weigh in on lie detection, or if you have other questions about true crime and forensics, seek me out at Quid Pro Quo Cast on Twitter, Facebook, Patreon, Gmail, at the.com, and wherever fine, Internets are sold. I am currently the most active on Twitter. Patreon patronage gets you a shout out on the podcast.
To that end, I'd like to thank my most recent patron, Leslie Levine from Shawville, Quebec. I'd also like to thank my newest iTunes and Stitcher's reviewers, Kay Coling, Violet Electric Pea, Bookshelf's 91, and Bossy Pants 16. April's Murder of the Month book pick is Tinseltown by William J. Mann.
Feel free to join us at the Murder of the Month page on Facebook for more information. Quid Pro Quo is produced from Stamp to Stearn by Me, Elle Stewart, and is available on iTunes, Stitcher, and Google Play. Please, if you do like the show, take a moment to subscribe, rate it, review it, and share it. This helps more than I can express.
Your support is greatly appreciated.