A polygraph is a set of sensors that records how your body reacts while you answer questions, and an examiner reads those reactions for signs of deception. The machine has no way to detect a lie itself. It picks up arousal, and fear, anger, or confusion can produce the same reactions as guilt. So the honest answer is that it works better than guessing in some settings, but nowhere near as reliably as its reputation suggests.1PubMed. Current status of forensic lie detection with the comparison question technique: An update of the 2003 National Academy of Sciences report on polygraph testing How far you can trust it depends heavily on why you are being tested, and screening job applicants is the use with the weakest scientific support.
What the Polygraph Machine Measures
Four body signals are recorded, each with its own sensor. Two stretchy bands around your chest and belly track your breathing as you inhale and exhale.2EuroPolygraph. What Is a Polygraph Test and Does It Really Work? A blood pressure cuff on your upper arm, or a sensor on your finger, tracks your pulse and blood pressure.3EuroPolygraph. What Is a Polygraph Test and Does It Really Work? Small plates on your fingertips measure sweating by how easily a tiny electrical current passes across your skin.4EuroPolygraph. What Is a Polygraph Test and Does It Really Work? All of these are controlled by the part of the nervous system that runs automatically, the same system that speeds your heart when you are startled.5Polygraph. Scientific Basis for Polygraph Testing
Being hooked up is more tedious than frightening. The cuff tightens and may feel uncomfortable for several minutes at a time. You are asked to sit still with your feet flat on the floor, because movement, deep sighs and coughs show up on the charts and can be flagged. Modern machines feed the readings into a computer instead of drawing them with ink pens.
Despite what TV suggests, the machine never beeps at a lie. No single reading marks one; the examiner interprets the charts afterward.
Examiners combine readings from several sensors and ask each key question more than once.6Polygraph. Scientific Basis for Polygraph Testing One spike on one signal could be a cough or a stray thought, so they look for a pattern that repeats. What the chart shows is a change from your own baseline, so a reaction means something only when compared with how you responded to other questions in the same session.
How a Polygraph Exam Works Step by Step
A typical exam starts with consent paperwork and a pretest interview, which often takes longer than the test itself. The examiner reviews every question with you beforehand, so there are no surprise questions. You may be asked about your health, medicines and sleep. Examiners often say the machine is very accurate, partly to raise the stakes. Anything you admit during this talk can be recorded and used, separately from the charts.
Next the sensors are attached and a practice round is run. Many examiners ask you to pick a number or card and deny it, partly so you get used to the sensors and the routine, and partly to show you the machine seems to work. That demonstration is not evidence of accuracy, since the examiner knows the answer in advance.
Most exams then use the comparison question method (also called the control question method). It mixes relevant questions, which are about the issue under investigation, such as the theft, with vaguer comparison questions.7PubMed. Effect of two types of control questions and two question formats on the outcomes of polygraph examinations A comparison question might be “Before age 21, did you ever take something that wasn’t yours?” or simply “Did you ever…?” with no time limit.8PubMed. Effect of two types of control questions and two question formats on the outcomes of polygraph examinations
The idea is that an innocent person will be more worried by the vague questions, which almost everyone has to answer with a small lie, while a guilty person reacts more to the questions about the crime. That assumption, that an honest person worries more about the comparison questions than the real ones, has been judged scientifically weak.9PubMed. Current status of forensic lie detection with the comparison question technique: An update of the 2003 National Academy of Sciences report on polygraph testing Examiners often steer you toward answering “no” to a comparison question, for example by suggesting that an honest person would never have done such a thing. If you answer “no” while privately remembering a past slip, the test treats that small lie as a baseline for how you react when being untruthful. The same question list is repeated several times.
The examiner, or a computer program, scores the size of your reactions to each kind of question across several rounds.10Polygraph. Scientific Basis for Polygraph Testing If you react more to the questions about the issue than to the comparison questions, the result leans toward deception; the reverse leans toward truth. The outcome is usually reported as deception indicated, no deception indicated, or no clear result. A result with no clear answer is common and often leads to a retest. A post-test talk follows, where the examiner may press you about reactions they saw.
How Accurate a Polygraph Really Is
The U.S. National Academy of Sciences found the evidence behind the standard polygraph method weak and mostly low quality.11PubMed. Current status of forensic lie detection with the comparison question technique: An update of the 2003 National Academy of Sciences report on polygraph testing It concluded that the industry’s claims of high accuracy were not backed up, and that the real error rate is unknown, even though the test beats chance.12PubMed. Current status of forensic lie detection with the comparison question technique: An update of the 2003 National Academy of Sciences report on polygraph testing A later review found that research quality had barely improved and that those conclusions still stand.13PubMed. Current status of forensic lie detection with the comparison question technique: An update of the 2003 National Academy of Sciences report on polygraph testing
Where the test is used matters. A U.S. Congress research office found meaningful evidence that the polygraph works only for investigating a specific crime.14U.S. Government Printing Office. Scientific Validity of Polygraph Testing: A Research Review and Evaluation It found no established scientific validity for screening employees or security clearance applicants.15U.S. Government Printing Office. Scientific Validity of Polygraph Testing: A Research Review and Evaluation A screening exam asks broad questions about things that may never have happened, so there is no single event to react to.
Optimistic numbers can mislead. In one study of real criminal cases, examiners who saw the case details judged 91% to 96% of truthful answers correctly.16National Institute of Justice. Study of the Validity of Polygraph Examinations in Criminal Investigation Examiners reading the charts without the case details judged truthful people correctly only 63% to 85% of the time.17National Institute of Justice. Study of the Validity of Polygraph Examinations in Criminal Investigation Knowing the case appears to inflate accuracy, and honest people are the ones most likely to be wrongly flagged.
Lab studies flatter the test for a similar reason. Many use volunteers who commit a mock crime and then deny it, and they have little to lose from being caught. A real suspect or job applicant may face prison or a lost career, which changes how the body reacts and makes the comparison unreliable. So a high figure from a lab study does not tell you how well the test would do on you.
Because the evidence is weak, a “deception indicated” result is better read as a reason for more questions than as a finding of lying. The test can help steer an investigation, for example toward a suspect worth a closer look, without establishing guilt. A pass does not prove innocence, since some liars do pass.
Why Honest People Fail and Liars Pass
Guilty people trained to bite their tongue, press their toes into the floor, or count backward by 7 during the comparison questions beat the test about half the time.18PubMed. Mental and physical countermeasures reduce the accuracy of polygraph tests The physical and mental tricks worked equally well, mainly by changing heart and blood pressure readings.19PubMed. Mental and physical countermeasures reduce the accuracy of polygraph tests Examiners and their equipment had trouble spotting them.20PubMed. Mental and physical countermeasures reduce the accuracy of polygraph tests The tricks work by raising your reaction to the comparison questions, so your reaction to the questions about the issue looks smaller by comparison.
Trying this yourself is risky. Examiners watch for unusual breathing and movement, and being caught using a trick is often treated as failing. The experiment showing tricks work involved trained people, not someone following online tips.
Health and medicines can distort results too. Conditions that affect the automatic nervous system, such as some heart rate and blood pressure disorders, can make results inaccurate.21PubMed Central. Beyond the Polygraph: Deception Detection and the Autonomic Nervous System Taking several medicines can also muddy the readings.22PubMed Central. Beyond the Polygraph: Deception Detection and the Autonomic Nervous System Yet in one experiment, common calming drugs such as diazepam and the blood pressure drug propranolol did not help guilty people pass a guilty-knowledge test.23PubMed. Use of Antianxiety Drugs as Countermeasures in the Detection of Guilty Knowledge
Language matters as well. Lying in a second language can raise sweating more, because of the added worry of speaking a non-native language, on top of the emotions that come with lying.24International Journal of Psychophysiology. Emotion and lying in a non-native language A test given in your weaker language may record extra arousal that has nothing to do with lying.
So tell the examiner beforehand about every medicine you take, including blood pressure and heart drugs, and any nervous system, heart or anxiety condition. Mention poor sleep or illness that day, and ask for the exam in the language you speak most comfortably.
Polygraphs for Jobs, Probation and Sex Offender Supervision
For police and federal jobs, the questions cover drug use, crimes and honesty on your application. What you admit in the interview often matters more than the charts. An applicant who admits past drug use or an undisclosed offence may be rejected on that admission alone, regardless of the result. In the U.S., most private employers are barred by federal law from requiring a polygraph.
Post-conviction testing is a different setting. Routine polygraph testing of convicted sex offenders is common in the U.S., is growing in England and Wales, and is rare elsewhere.25PubMed. Post-conviction polygraph testing of sex offenders Probation services in England and Wales increasingly use it to draw admissions from released offenders, despite scientific criticism.26CrossRef API / The Journal of Criminal Law. Not ‘Very English’ – on the Use of the Polygraph by the Penal System in England and Wales Its main value is that people disclose more past offending and rule-breaking when they are tested.27PubMed. The (f)utility of post-conviction polygraph testing There is not enough evidence on its accuracy or on whether it deters reoffending.28PubMed. The (f)utility of post-conviction polygraph testing
People who are tested describe real costs. In one survey of offenders in treatment, about 7% of tests wrongly accused the person of deception, by their own account.29PubMed. Post-conviction sex offender polygraph examination: client-reported perceptions of utility and accuracy About 5% said they had admitted to things they hadn’t done after being wrongly accused.30PubMed. Post-conviction sex offender polygraph examination: client-reported perceptions of utility and accuracy
If you are on probation and told to take one, refusing can count as a supervision violation, and admissions you make can trigger sanctions. Ask your lawyer or officer how results are used in your case. Do not confess to something untrue to end the pressure.
Is a Polygraph Result Allowed in Court
Polygraph results are inadmissible in Canada and in many U.S. jurisdictions.31Law and Human Behavior. The admissibility of polygraph evidence in court: Some empirical findings Courts doubt their scientific reliability, the same doubts raised by the major reviews. Statements and confessions made during a polygraph session can still be used as evidence. The practical risk of a police polygraph lies in what you say, not in how the machine scores you.
That explains why police use polygraphs even when courts won’t accept them. They serve as an interrogation tool to get people talking, and a failed result may be used to pressure a suspect.
If police ask you to take one, it is usually voluntary when you are a suspect. Passing does not guarantee that police drop the case. Talk to a lawyer before agreeing. Some U.S. states allow results if both sides agree in advance.
Newer Lie Detection Methods
One alternative is the Concealed Information Test, also called the guilty-knowledge test. It asks whether you recognize details only the culprit would know, such as the item that was stolen, rather than whether you lied.32FBI: Law Enforcement Bulletin. The Concealed Information Test: An Alternative to the Traditional Polygraph It may work better for people who feel little fear, such as psychopathic offenders, who don’t react the way the standard test expects.33FBI: Law Enforcement Bulletin. The Concealed Information Test: An Alternative to the Traditional Polygraph
It differs from the standard exam in what it asks. It does not ask whether you did something; it shows you a set of options, such as several items, and checks whether one stands out to you. An innocent person has no reason to recognize the real item, so their reactions should look the same across the options. It requires that investigators know details that have not been made public, so it can only be used in some cases.
Brain-based tests are another route. Brain scans in a scanner have told lies from truths in healthy volunteers more than 75% of the time.34PubMed Central. Using Brain Imaging for Lie Detection: Where Science, Law and Research Policy Collide In one mock-theft lab experiment, scans identified the stolen object for about 9 in 10 people.35PubMed. Detecting deception using functional magnetic resonance imaging Brainwave tests look for a recognition signal when someone sees crime-related details.36PubMed Central. Brain fingerprinting: a comprehensive tutorial review of detection of concealed information with event-related brain potentials One such method reported perfect accuracy in its developer’s own field studies, a claim that needs independent checking. Another brainwave method correctly classified 86% of guilty and innocent people in the lab.37PubMed. A new approach for EEG feature extraction in P300-based lie detection
Eyes and thinking effort have been tried as well. Pupils tend to widen a little more when people lie.38PubMed Central. Truth and lies in your eyes: Pupil dilation of White participants in truthful and deceptive responses to White and Black partners Making questions mentally harder can expose liars better than making them emotional.39한국심리학회지: 법. The Difference in Pupil Size Responding to Cognitive Load and Emotional Arousal Questions between Guilty and Innocent Groups In one experiment, how slowly people answered unexpected questions predicted lying better than their eyes did, with up to 84% accuracy.40Legal and Criminological Psychology. The lie in the eyes: Ocular indicators and machine‐learning models with a cognitive load paradigm to detect lying Newer AI scoring programs outperformed older computer scoring in small tests.41PubMed Central. Development of a Deep-Learning-Based Computerized Scoring Algorithm These improvements are early and mostly come from lab studies. None of these tests is routinely used for job or probation testing, and brain scans are expensive and not accepted as courtroom evidence.