How a polygraph works: breathing, pulse and skin sensors explained

How a polygraph works: breathing, pulse and skin sensors explained simply

Before every test people ask me: “What exactly are you going to measure?” The answer is simple: a polygraph does not read minds and does not “see lies”. It accurately records how the body reacts to questions. The conclusion is drawn by the examiner, who compares these reactions according to strict rules.

Need an actual test rather than theory? Take the VerifEye online test — from 2,500 UAH (≈ €50), with no office visit, or calculate the cost of a polygraph test.

In this article I, polygraph examiner Andriy Maryniak, explain which sensors are attached during the test, what each of them shows and how a conclusion is reached from the charts.

Which sensors a modern polygraph uses

The American Polygraph Association (APA) requires in its standards that the instrument records at least:

• breathing — two sensors: on the chest and on the abdomen;
• electrodermal activity — how the skin’s conductance changes;
• cardiovascular activity — pulse and relative blood pressure;
• body movement — a separate motion sensor; under APA standards it has been mandatory since 2012.

Many instruments also have a photoplethysmograph — a finger sensor that shows how the blood volume in the vessels changes. It is optional but gives extra information.

All sensors are placed on the outside of the body and pierce nothing. The skin sensor passes a micro-current that a person cannot feel. The test is safe for health.

Breathing sensors (pneumographs)

Two corrugated tubes go around the upper chest and the abdomen. As a person breathes, their circumference changes — this is how the instrument records the rate, depth and rhythm of breathing.

When a question matters to a person, breathing often changes involuntarily: it becomes shallow, slows down or stops for a few seconds. There are two sensors — chest and abdominal — so the examiner can see where the person breathes naturally and where they are trying to control their breathing. This is the channel people can influence most easily on purpose, so it is never evaluated on its own.

Skin sensor (electrodermal activity, EDA)

Two small plates are attached to the fingertips of two fingers — usually the index and ring fingers of the hand a person uses less. A very weak current passes through the skin, and the instrument measures how its conductance changes. It depends on the sweat glands, which are controlled by the nervous system, not by a person’s will.

Research shows this is the most informative channel: it contributes the most to the final score. Older literature calls it GSR — the galvanic skin response.

Cardio channel: pulse and blood pressure

A regular medical cuff, like the one on a blood-pressure monitor, is put on the upper arm of the other hand, but it is inflated much less than when a doctor measures blood pressure — just enough to record the pulse wave.

The chart shows relative changes in blood pressure and the pulse. A reaction to a meaningful question usually looks like a short rise in the blood-pressure “baseline”. Contrary to popular belief, a “faster heartbeat” carries little weight in the conclusion — what matters most is the change in blood pressure.

Photoplethysmograph (finger sensor)

A small optical clip-on sensor on the finger records how the blood volume in the vessels changes. When a person tenses up, the peripheral vessels narrow and the signal amplitude drops. Some techniques award separate points for this; others use the channel as additional confirmation.

Motion sensor

Flat sensor pads are placed on the seat, under the armrests and under the feet. They pick up even small movements: tensing muscles, shifting, pressing the toes.

Why is it needed? Movements distort the other charts, and segments with movement are not scored. Most importantly, it catches attempts to “adjust” the reactions: tensing muscles, biting the tongue, pressing the toes. Research shows that without this sensor examiners spot such attempts poorly, and with it much better: in one study a chair with sensors detected 85% of physical countermeasures. That is why, since 2012, APA standards have required a motion sensor in every test.

How the charts turn into a conclusion

The test consists of several recordings of 4–7 minutes each. Usually there are three; if the result is unclear, two more are added. Each recording contains three types of questions:

• relevant — about the event being checked (“Did you take money from the till?”);
• comparison — questions that almost anyone who is nervous about the test itself reacts to;
• neutral — to let the body return to a calm state.

The examiner compares which questions produce stronger reactions: the relevant or the comparison ones. A stronger reaction to a relevant question gives a minus score, to a comparison question a plus score. Scores are assigned on standardised scales (for example, ESS), and modern software also calculates the result with an algorithm. The conclusion is based not on a single “spike” on a chart but on the total score across all recordings and all channels.

We always agree the questions in advance: the person knows every question before the recording starts. There are no surprises in the test.

What prevents an accurate result

• Fatigue and lack of sleep — reactions become weak and unstable.
• Alcohol, sedatives, energy drinks — they change physiology; any medication must be mentioned in advance.
• Acute illness, pain, high temperature — in such a state the test is rescheduled.
• Countermeasures — controlled breathing, pain, muscle tension. The examiner looks for them on the charts and on the motion sensor.
• Creams, lotions, a cold room — they distort the skin-sensor signal, so hands should be clean and warm before the test.

How EyeDetect, VerifEye and StimulTest work

These technologies need no sensors on the body:

• EyeDetect — the station’s camera tracks the eyes: pupil size, gaze movements, blinking, how a person reads the statements, as well as the speed and accuracy of the answers. The result is calculated by an algorithm. It is based not on emotional reaction but on cognitive load: lying is harder than telling the truth.
• VerifEye — a similar principle, but the test is taken online in a smartphone app: the camera records the eyes, and the person answers the questions out loud.
• StimulTest — measures reaction time: the person presses a button, and the system calculates the difference in milliseconds.

A comparison of all four methods — format, duration, accuracy and price — is in the table on the prices page.

How accurate is it

According to the American Polygraph Association meta-analysis (2011), polygraph accuracy is 89% for a single specific incident and 85% for multi-issue screening. It is not 100%, so the examiner’s conclusion is an important but not the only basis for decisions.

Frequently asked questions (FAQ)

Does a polygraph test hurt?

No. The sensors only rest against the skin and pierce nothing; the skin sensor’s micro-current cannot be felt. The only sensation is the moderate pressure of the cuff on the arm during a recording.

Is a polygraph harmful to health?

No. The instrument only records the body’s signals. During the pre-test interview the examiner asks about health: in case of acute illness, severe fatigue or pregnancy the test is rescheduled or not carried out.

Can the sensors be fooled?

The examiner looks for attempts to control breathing, tense muscles or cause pain on the charts, and the motion sensor catches most physical attempts. Besides, such “tricks” only harm an innocent person: they increase the chance of looking deceptive. The best strategy is to answer truthfully and calmly.

How long does a person sit with the sensors on?

One recording takes 4–7 minutes, usually three to five of them. The whole examination, with the interview and agreeing the questions, takes 1.5 to 3 hours: APA standards do not allow a full examination to be shorter than 90 minutes.

Does nervousness affect the result?

Everyone is nervous, and the technique accounts for it: it compares not “calm versus nervous” but the strength of reactions to different types of questions.

About the author: Andriy Maryniak — polygraph examiner in Lviv
How the test works: polygraph test — what to know

Article by polygraph examiner Andriy Maryniak, in practice since 2012.