CIT Test vs Polygraph

The Short Answer

They are not two versions of the same thing. They rest on different ideas about what gives a person away.

A polygraph examination, in the form most people have seen, asks the examinee direct questions about what they did and records how their body reacts while they answer. The reasoning is that a person who is lying about something important to them will be more aroused when asked about it than when asked about something else.

A Concealed Information Test does not ask what the person did. It shows them a set of items, one of which is a true detail of the event, and measures whether that one item is processed differently from the rest. The reasoning is that a person recognises what they know, whether or not they want to.

The first looks for a reaction to being accused. The second looks for a reaction to something being familiar. Almost every practical difference between the two follows from that.

How a Polygraph Examination Works

A fair description, because a comparison is only useful if both sides are described accurately.

A polygraph is an instrument that records several bodily signals at once. Typically these are breathing, through bands around the chest and abdomen; skin conductance, through sensors on the fingers, which rises with sweat gland activity; and cardiovascular activity, through a blood pressure cuff. Some examiners add a movement sensor in the seat.

The instrument does not detect lies. It records arousal. What turns it into a test of deception is the questioning technique laid on top. The most widely used is the comparison question technique, sometimes called the control question test. The examiner spends a long pre-test interview going through the questions with the examinee. There are relevant questions, about the matter itself ("Did you take the money from the office on the fourth of March?"), and comparison questions, which are deliberately broad and uncomfortable ("Before this year, did you ever take something that did not belong to you?"). The examinee is expected to answer no to both kinds.

The theory is that an innocent person, who can answer the relevant questions truthfully, will be more troubled by the comparison questions, where they are probably shading the truth or at least unsure. A guilty person will be more troubled by the relevant questions, because those are the ones that threaten them. The examiner scores the charts by comparing the size of the reactions to the two kinds of question.

The technique's strength is that it can be pointed at a direct question. If all you have is "did you or did you not", it is the approach that fits.

How a CIT Works

The outline; the detail is on the how CIT testing works page.

Before the test, the examiner and the person arranging it choose details of the event that the tested person has not been told. For each detail, a list is written: the true item, called the probe, and five plausible false ones, called irrelevants. A further item, the target, gives the tested person a simple key-press task.

The items appear one at a time on a screen, each shown twenty to forty times in shuffled order, while an EEG cap records the brain's electrical activity. The analysis averages the response to each kind of item and compares the probe with the irrelevants. A reliably larger P300 to the probe indicates that the person's brain treats it as familiar.

The CIT's strength is that its logic is simple and its comparison is like for like. Every item on the list is the same kind of thing, presented in the same way. The only difference between the probe and the irrelevants is whether the person knows it is real.

Side by Side

The main differences in one table.

The Nervousness Problem

This is the question people ask most: "I get anxious. Won't that make me fail?"

It is a reasonable worry about any test built on arousal. Being asked "did you steal from your employer?" is unpleasant whether or not you did. The comparison question technique tries to deal with this by giving the innocent examinee something even more unsettling to react to, but that depends on the comparison questions landing as intended for that particular person, and critics of the technique have long argued that this cannot be guaranteed.

A CIT handles the problem by design. Nervousness affects every item equally. A frightened person is just as frightened when "red folder" appears as when "blue cash tin" does. Since the result comes from the difference between items, general anxiety does not push it either way. And because the P300 reflects attention and recognition instead of arousal, it is a different kind of signal from the ones a polygraph records.

That does not make the CIT safe from error for a truthful person. Its equivalent weakness is leaked knowledge. If an innocent person has been told the true detail, they will recognise it. The risk has moved, not vanished: with a polygraph the danger is an innocent person who is upset; with a CIT it is an innocent person who is informed.

What the Research Says About Each

Both methods have been studied for decades. Neither comes out as a simple truth machine.

Polygraph

The most cited independent review is the 2003 report of the US National Research Council, The Polygraph and Lie Detection. Its committee concluded that, in populations of examinees untrained in countermeasures, specific-incident polygraph tests can discriminate lying from truth-telling at rates well above chance, though well below perfection. It also judged the quality of much of the research to be low, found the theoretical basis weak, and warned that accuracy in real conditions is likely to be lower than in the studies. The committee was particularly critical of using the polygraph for screening, where there is no specific incident.

CIT

The CIT has a stronger standing among academic psychologists, largely because its logic is transparent and its comparison is controlled. The 2014 meta-analysis by Meijer and colleagues pooled the published studies and found that all four measures examined distinguished people with knowledge from people without, with effect sizes of 1.89 for the P300, 1.55 for skin conductance, 1.11 for respiration and 0.89 for heart rate. The authors noted that the P300's advantage held in studies using personal items and not in mock-crime studies.

Laboratory studies of the P300 CIT commonly report detection of 80 to 95 per cent of people holding the tested information. Those are laboratory figures and are not results from our own testing. The CIT's weak point in the research record is the opposite of the polygraph's: it has good laboratory evidence and relatively little field validation outside Japan, whereas the polygraph has enormous field use and contested laboratory evidence.

Countermeasures

Both can be attacked. The National Research Council report noted that countermeasures against the polygraph are a serious concern. For the P300 CIT, Rosenfeld and colleagues showed in 2004 that simple trained countermeasures could defeat the three-stimulus protocol in a substantial share of participants, and they later developed the Complex Trial Protocol to resist this. We use the three-stimulus protocol and do not claim it cannot be beaten.

Polygraph and the Law in the UK

Where the polygraph is officially used, and what the statute says about its results.

In England and Wales the polygraph has a defined role in the supervision of certain offenders released on licence. The Offender Management Act 2007 allows a polygraph condition to be included in the licence of people convicted of specified sexual offences. Later legislation extended polygraph conditions to some domestic abuse offenders, initially as a pilot under the Domestic Abuse Act 2021, and to some terrorism offenders. The examinations are carried out for the probation service and are used to help manage risk and check compliance with licence conditions.

The same Act limits what can be done with the results. Section 30 provides that statements made in a polygraph session, and the physiological reactions recorded during it, may not be used in evidence against the released person in any proceedings for an offence. In other words, even where Parliament has authorised polygraph testing, it has kept the results out of criminal proceedings against the person tested.

There is no equivalent statutory scheme for the CIT in the UK. It is a private, voluntary test. We make no claim that a CIT result is admissible in any court or tribunal, and nobody should book one expecting it to be. If you need evidence for proceedings, ask a solicitor what the court will accept.

Which One Fits Your Situation?

Start from what you actually have.

A CIT fits when

• You know specific details of what happened, and the other person has not been told them.
• You want a test that does not put accusing questions to the person.
• The person is anxious, and you are concerned that a test based on arousal would be unfair to them.
• You want a result computed from a recording that is retained and can be re-analysed.

A CIT does not fit when

• All you have is a general suspicion with no specific details.
• Every detail has already been put to the person.
• The question is about intention, feeling or opinion, such as "do you still love me?", which has no detail to recognise.

In the second group, a comparison question polygraph is the type of test designed for a direct question, with the limitations described above. We do not provide polygraph examinations and do not recommend particular examiners. If you are considering one, ask the examiner the same things you should ask us: what exactly is measured, what technique is used, what independent research supports it, and what happens when the result is inconclusive.

Where there are no known details but there is a short list of possibilities, a searching CIT sits between the two. It keeps the CIT's like-for-like comparison but gives weaker results, and that page sets out the trade-off.

Where the Two Tests Came From

A short history explains why they are so different.

The polygraph grew out of early twentieth-century work on recording blood pressure and breathing during questioning. By the middle of the century it had become an instrument used by police forces and employers, mainly in the United States, and the comparison question technique was developed to give examiners a structured way of interpreting the charts. It was a practitioner's tool first, and the research followed.

The Concealed Information Test took the opposite route. David Lykken, a psychologist at the University of Minnesota, was critical of polygraph questioning and proposed an alternative in a 1959 paper. His point was that an examiner should not try to detect lying at all. An examiner should test for knowledge, using multiple-choice questions to which only an involved person would know the answers. He called it the Guilty Knowledge Test and demonstrated it with skin conductance.

In the late 1980s and early 1990s, researchers including Peter Rosenfeld at Northwestern University, and Lawrence Farwell and Emanuel Donchin at the University of Illinois, replaced the bodily measure with the P300. The brain response had a clearer theoretical link to recognition than sweat gland activity did, and it could be averaged over many presentations. That line of work produced the three-stimulus protocol and, later, the Complex Trial Protocol.

So one test began in practice and has been examined by researchers ever since. The other began as a researcher's proposal and has been adopted in practice in only a few places, the most notable being Japan.

Misconceptions About Both Tests

A few beliefs that get in the way of a sensible choice.

• "A lie detector detects lies." Neither test does. A polygraph records arousal and infers deception from a comparison. A CIT records a brain response and infers recognition from a comparison.
• "A brain-based test must be more certain than a body-based one." Being recorded from the head does not make a result infallible. A CIT is only as good as its item lists and the secrecy of its details.
• "A pass proves innocence." No test proves a negative. A flat CIT result shows no sign of recognition for the details tested. A truthful polygraph result shows no stronger reaction to the relevant questions. Both leave room for error.
• "An inconclusive result is a hidden fail." It is not. In a CIT, an unusable block means the recording did not meet the standard needed to say anything.
• "Either test can be used on someone without their knowledge." Both require the person's active cooperation and consent.

How the Two Feel to Take

The experience differs as much as the method.

A polygraph examination is an interview. It commonly lasts two hours or more, most of it spent talking with the examiner before the instruments are attached. The examinee is asked about the allegation directly and repeatedly. For some people that is exactly what they want: a chance to deny it formally. Others find it confrontational.

A CIT is a task. After consent and set-up, the person sits in front of a screen and presses one of two keys as words appear. Nobody asks them whether they did anything. Each block lasts five to ten minutes, with rests between, and the whole appointment is usually 60 to 90 minutes. The most common comment afterwards is that it was boring. From a measurement point of view, that is ideal.

One practical consequence is worth knowing. Because a CIT does not involve an interview about the allegation, the examiner does not form a view of the person's demeanour and does not report one. The result is what the recording shows.

Where We Stand

We offer one of these tests, so here is our interest declared and our reasoning.

Deception Detection UK provides P300 EEG Concealed Information Tests using a four-channel BrainBit cap, with a procedure based on Rosenfeld's published P300 CIT protocols. We chose the CIT because its logic can be explained in a paragraph, its comparison is controlled, its result is calculated and can be checked, and it carries a built-in test of its own validity in the target check.

We also accept its limits. It cannot answer a bare accusation. It depends on details staying private. Its field validation is limited. Where a case does not fit, we say so. You can read the requirements on the who qualifies page, or return to the CIT testing overview.