ITV’s hugely successful and long-running Jeremy Kyle Show has been cancelled following the suspected suicide of a guest who failed a ‘lie detector’ test. The myth of the high accuracy of ‘lie detectors’ must be dispelled.
Of course, the Jeremy Kyle Show isn’t the first programme to use a lie detector on guests. The lie detector test, or polygraph test as it is more correctly known has become a staple of both US and UK daytime talk-show programming. But, despite the claims of the hosts of these programmes, the polygraph is far and away from reliable.
In fact, simple knowledge of what it is that the polygraph actually measures should alert most people to the fact that the results are hardly indicative of anything other than the subject being alive.
Couple this with is the fact that the results gained from a polygraph session are ultimately examined by a human being who will determine what is a ‘lie’ and what is not. Thus meaning that what is presented as objective science is actually very subjective–based on the biases and prejudices of a single person.
The myth that the polygraph is so accurate in the public consciousness and its use on daytime TV might actually be responsible for the death of a man in the UK.
Steve Dymond died of a suspected overdose a week after failing a polygraph test in the studio of the incredibly popular daytime programme ‘the Jeremy Kyle Show’.
Dymond had agreed to come on the show–now cancelled by ITV–to prove to his fiancée that he had not been cheating on her.
Sadly, in my personal opinion, this was an incident that has been waiting to happen. Much attention has been cast on the ethical issues surrounding the vetting of guests on shows such as ‘the Jeremy Kyle Show’, but the deeper problem may lie with the way the polygraph is used by these programmes and the myth that has built up around them.
The ‘lie detector’ as dramtic denouncement
The lie detector test fulfils an important function on daytime talk shows. It can be understood as a factor of the narrative these programmes seek to weave several times an episode.
In the short 15-30 segments in which a scenario is introduced, the alternative side is presented and the affair descends, generally, into a screaming match between two parties–the audience is offered little in terms of resolution.
The ‘lie detector’ is offered as such resolution, but for it to be valid, arguments against its validity have to be pre quelled. That’s how the myth about “99% accuracy” and similar claims of accuracy has built up.
In the quest to provide audiences with satisfying conclusions to the segments in the show and prevent lingering niggles to carry forward as the next group of unfortunate people are paraded onto the stage–the producers of such television have elevated a piece of border-line pseudoscience to infallibility.
What does a polygraph actually measure?
The key to understanding why a polygraph isn’t actually a conclusive ‘lie detector’ lurks in the knowledge of what factors it measures, and how such conditions arise.
The polygraph measures proxies such as heart-rate, breathing and galvanic skin response–the amount of electricity conducted by the skin as a result of sweating. The idea is that if a person is lying, their heart rate will increase, their breathing will become shallower and they will begin to sweat.
The polygraph operator will ask a series of ‘dummy’ questions in an attempt to get a baseline reading of these physical conditions in a subject. Then they move on to questions related to the actual matter to be investigated–in this tragic case, infidelity.
The theory goes that significant deviations from these baseline readings will indicate attempted deception as a result of the stress that places on the body.
The flaws here should be pretty self-evident. Crucially, the act of lying isn’t the only thing that causes stress!
So a ‘lie detector’ doesn’t detect ‘lies’ or ‘truth’–it detects stress response.
There is very probably preexisting stress associated with questions like “have you cheated on your partner?” if a subject knows their relationship hinges on this question and, crucially, they have been anticipating it. Thus their stress response is already likely to be elevated as soon as the question is asked.
The British Psychological Society agree. In a 2004 report, they called it ‘naive’ to assume that the effect of this type of questioning would not lead to a stress response.
From the report: “when an innocent man, suspected of murdering his beloved wife, is asked questions about his wife in a polygraph test, the memory of his late wife might reawaken his strong feelings about her.
“Simple fear of being disbelieved could make one more anxious when answering the serious questions.”
A further underlying issue with polygraphs and their use in stressful situations is how hard it is to conduct an accurate test. A lab-based experiment with test subjects is unlikely to generate the same kind of conditions that guests on Kyle’s show experience.
Thus, there’s no real way to test if the polygraph actually works in situations that ‘the Jeremy Kyle show’ and other talk shows create.
Dr Jamie Horder, a neuroscientist and editor at the journal Nature Communications, tells the Telegraph: “In any kind of experiment you have to know what’s called the ‘ground truth’, or in this case whether someone is lying or not. You can do it in a lab, because you can say to half the people ‘I want you to tell the truth’, and the other half ‘I want you to lie’.
“But in that case, there’s very little consequence involved, the levels of stress in the test are much lower than if someone’s facing real consequences. And that’s really important because the test is all about stress. It’s essentially testing the stress response.”
The other issue with lie detectors is that a better, more confident liar is less likely to experience stress and therefore produce an above baseline response than an extremely nervous innocent person.
A 1986 study published in the Lancet seems to support this conclusion, determining that an ‘innocent’ verdict from a polygraph test was far more likely to be reliable than a ‘guilty’ verdict.
The two measures of ‘success’ in polygraph testing
Part of the problem with the myth of polygraph accuracy is the idea that there is one single measure of how successful such a test is.
In medical testing, there are two measures of accuracy–sensitivity and specificity. Sensitivity is the extent to which actual positives are not overlooked–in our lie detector scenario this would be the number of times a lie is correctly identified as such.
The aforementioned Lancet review gave the polygraph a sensitivity rate of around 75%. This means for every 100 false statements, 25 may be incorrectly identified as ‘true’. Thus we say the lie detector doesn’t throw up a lot of false negatives (positive being a lie and negative the truth in this case).
The problem lies in the polygraph’s specificity, which the 1986 Lancet review found was around 50% – 60%. That means that for every 100 true statements made, 40 – 50 of them could potentially be identified as ‘false’. Thus the lie detector has a very high chance, according to the Lancet,
of throwing up a false positive.
That means that if a polygraph test says someone’s truthful, you can be much more confident that it’s right than if it says someone’s lying.
That’s a big problem for a system that claims to be detecting lies.
Why cancelling the Jeremy Kyle Show is not enough
With ‘the Jeremy Kyle Show’ confined to the dustbin of television history, all that it does is leave a gap in ITV’s schedule to be filled by another talk-show with a similar tone.
Before Kyle took that spot, it was occupied by Vanessa Feltz in a similar style of programme. When that show was hit by scandal, ITV simply switched it out with a programme with the same template format–hosted by Trisha Goddard–eventually superceded by Kyle.
The changes have been little more than cosmetic. Do have any reason to suspect ITV won’t just do the same again?
Any show that rises up after ‘Kyle’ will likely resort to lie detector tests again for a quick and easy denouncement. This will likely be prefaced with an on-screen caption warning of disputes over the accuracy, but the show will just plug into and feed off the same misinformation about lie detectors shown by its predecessors.
This myth leaves no avenue for doubt–in the mind of the misinformed, the subject, their friends and family they are undoubtedly guilty–even if the subject knows they are innocent.
The effect of that stigma on a person with mental health issues took lead to tragic circumstances. That means that OFCOM has to act and regulate the use of the polygraph in talk shows for what is ultimately for the entertainment of others. Hopefully, that will lead to other countries follow suit.
Otherwise, this myth of infallibility will likely lead some other unfortunate person to harm themselves or someone else.
That’s the truth.