Polygraph Accuracy, Base Rates and Predictive Value Explained

When people ask “How accurate is a polygraph?” they usually expect a single number. The question is understandable, but the answer is more involved than any headline figure can convey. A reported accuracy rate only becomes meaningful when it is placed in the context of the situation being investigated — and that context depends on something called the base rate.

Meta-analytic reviews of polygraph techniques — including the American Polygraph Association’s meta-analyses and the National Research Council’s 2003 review — report average accuracy figures, sometimes in the range of 80 to 90 per cent for well-validated techniques. These figures reflect sensitivity and specificity calculated across laboratory and field studies. They describe how well the technique distinguishes between deceptive and truthful responses under research conditions.

What they do not describe, on their own, is the probability that a specific person labelled as deceptive is actually lying, or that a person labelled as truthful is actually telling the truth. To understand those probabilities requires an additional piece of information: the base rate — the proportion of people in the relevant population who are actually engaging in the behaviour being tested for.

What Is a Base Rate?

A base rate is the underlying prevalence of a condition or behaviour in a given population. In medical testing, the base rate is the proportion of people in the tested group who actually have the disease. In polygraph testing, the base rate is the proportion of examinees who are genuinely being deceptive about the issue under examination.

Base rates vary enormously depending on context:

• Event-specific examinations — where someone is tested on a specific allegation, such as infidelity, theft, or a denial of a particular act — may have a moderate to high base rate of deception, because there is already reason to suspect the behaviour.
• Screening examinations — such as pre-employment vetting, periodic security screening, or post-conviction monitoring — typically have a lower base rate, because most people in the screened population are compliant.

The base rate is rarely known with certainty. It is estimated from the available evidence — the circumstances of the case, the referral context, known prevalence data, and professional judgement. But even an approximate base rate has a profound effect on how a test result should be interpreted.

Accuracy Is Not the Same as Predictive Value

This is the central point of the article, and it is a point that is widely misunderstood.

When a meta-analysis reports that a polygraph technique has, say, 90 per cent sensitivity (correctly identifying deceptive individuals) and 85 per cent specificity (correctly identifying truthful individuals), those figures describe the technical performance of the test. They answer the question: “Given that a person is deceptive, how likely is the test to detect it?” and “Given that a person is truthful, how likely is the test to confirm it?”

They do not answer the question most people actually want answered: “Given that the test says this person is deceptive, how likely is it that they actually are?”

That question is answered by the positive predictive value (PPV), and its companion, the negative predictive value (NPV). These depend not only on the test’s sensitivity and specificity but also on the base rate of deception in the population being tested.

Sensitivity, Specificity, PPV and NPV Explained

The following terms are standard in diagnostic testing across medicine, forensic science, and psychophysiology. Understanding them is essential for anyone interpreting polygraph results in a professional context.

• Sensitivity — the proportion of genuinely deceptive individuals whom the test correctly identifies as deceptive (the true positive rate).
• Specificity — the proportion of genuinely truthful individuals whom the test correctly identifies as truthful (the true negative rate).
• Positive Predictive Value (PPV) — given a positive result (the test indicates deception), the probability that the person is actually deceptive.
• Negative Predictive Value (NPV) — given a negative result (the test indicates truthfulness), the probability that the person is actually truthful.

Sensitivity and specificity are properties of the test. PPV and NPV depend on the test and the base rate. This distinction is crucial.

Why Low Base Rates Create False Positive Problems

Consider a screening scenario in which the base rate of the behaviour being tested for is low — say 5 per cent. If 1,000 people are screened, 50 are genuinely deceptive and 950 are truthful.

Using a test with 90 per cent sensitivity and 85 per cent specificity:

In this scenario, despite the test having respectable sensitivity and specificity, only about one in four positive results correctly identifies a genuinely deceptive person. The remaining three out of four are false positives — truthful people incorrectly labelled as deceptive.

This is not a failure of the test. It is the predictable mathematical consequence of applying any imperfect diagnostic procedure to a population in which the target behaviour is uncommon. The same arithmetic applies to medical screening tests, airport security checks, and many other diagnostic contexts.

Why High Base Rates Create a Different Problem

The reverse applies when the base rate is high. In a scenario where 80 per cent of examinees are genuinely deceptive — which might arise, for instance, in a post-conviction therapeutic context where the specific issue is known to be common — the PPV will be very high. Almost all positive results will be correct.

However, in that same scenario, the negative predictive value falls. A “passed” result becomes less reassuring, because the high base rate means a significant proportion of people who pass the test may in fact be deceptive but were not detected.

The practical implication is that a passed polygraph does not, by itself, guarantee truthfulness — just as a failed polygraph does not, by itself, prove deception. The weight of any result depends on the context in which it was obtained.

Event-Specific Testing and Screening Are Not the Same

This base-rate analysis has important consequences for the distinction between event-specific polygraph examinations and screening examinations.

An event-specific examination addresses a defined allegation or issue — for example, “Did you take money from the safe?” or “Did you have sexual contact with the person named?” In such cases, there is usually a reason for the test: a complaint, a suspicion, or a disclosure. The base rate of deception is typically moderate to high, because the referral itself was prompted by evidence or concern.

A screening examination is used to assess compliance or risk across a broader population — for example, pre-employment screening, periodic security vetting, or post-conviction sex offender monitoring. The base rate in screening populations is often lower, because most people in the group are compliant most of the time.

When interpreting results, professionals should consider which category the examination falls into, and what the likely base rate is. A positive result from an event-specific examination with a moderate base rate carries more predictive weight than the same positive result from a low-base-rate screening context.

The National Research Council’s Warning About Low Base-Rate Screening

The National Research Council (NRC) of the United States addressed this issue directly in its 2003 report, The Polygraph and Lie Detection. The report noted that even if polygraph accuracy were relatively high, the use of polygraph screening in populations where the base rate of the target behaviour is very low would inevitably produce a large number of false positives relative to true positives.

“The inherent ambiguity of polygraph testing makes its use in security screening problematic... The proportion of false positives will be high, even if the rate of false positive errors is low.”

This observation was not a rejection of polygraph science. It was a recognition that any diagnostic tool — however well-validated — faces predictive limitations when the target behaviour is rare. The NRC was applying the same reasoning that epidemiologists use when evaluating population-level screening programmes in medicine.

For professionals instructing or relying upon polygraph examinations, the NRC’s analysis is an important reminder that the context of use is as important as the accuracy of the test.

Polygraph Results as Information Gain

A more productive way to understand polygraph results is in terms of information gain — the extent to which the examination result shifts the estimated probability of deception from the prior (base rate) to a revised posterior probability.

In Bayesian terms, even a test with imperfect accuracy can provide substantial information gain in the right context. A positive result in a moderate-base-rate event-specific examination may shift the probability of deception from, say, 50 per cent to over 90 per cent. A negative result in the same context may shift the probability of deception from 50 per cent to under 15 per cent.

By contrast, the same test applied in a very low base-rate screening context may shift the probability from 2 per cent to 12 per cent — still meaningful as an indicator for further investigation, but far from conclusive. This framework — treating polygraph results as probabilistic evidence rather than binary verdicts — is consistent with the approach described in Raymond Nelson’s work on prior probabilities in polygraph screening.

Dr Keith Ashcroft has discussed the practical application of Bayesian reasoning to polygraph practice in a separate article on choosing the correct prior probability.

Practical Implications for Lawyers, Therapists, Investigators and Private Clients

Understanding base rates and predictive value has direct practical relevance for anyone who instructs, receives, or relies upon polygraph examination results.

For Solicitors and Barristers

A polygraph result is not a verdict. Legal professionals should understand that the predictive value of a result depends on the circumstances of the case. A positive result in an event-specific examination may carry substantial evidential weight as structured intelligence; the same positive result in a low-base-rate screening context carries less weight.

For Therapists and Clinical Professionals

In therapeutic polygraph contexts, where examination results contribute to disclosure, treatment planning, and risk assessment, the base rate of specific behaviours may be estimated from clinical knowledge. Results should be discussed with appropriate caution, especially where the base rate is uncertain.

For Investigators and Risk Managers

In security, corporate, and investigative contexts, the value of a screening result depends on the base rate of the behaviour being screened for. Where the base rate is low, a positive result should be treated as a reason for further investigation, not as proof of guilt.

For Private Clients

Individuals seeking a polygraph in connection with infidelity, false allegations, or other personal matters should understand that a polygraph result is a piece of structured evidence, not a guarantee. Its value depends on the quality of the examination, the specificity of the questions, and the context of the concern.

Responsible Interpretation of Polygraph Results

A responsible approach to polygraph interpretation requires:

• Acknowledging that reported accuracy figures from meta-analyses describe test performance under research conditions, not the probability that a specific result is correct in a given case.
• Estimating, where possible, the base rate of the behaviour being examined — and adjusting expectations about predictive value accordingly.
• Treating positive and negative results as probabilistic indicators, not binary truths.
• Recognising that a polygraph result is one element within a broader process of assessment, investigation, or risk management.
• Ensuring that question formulation is specific, testable, and appropriate to the context.
• Reporting results with appropriate caveats and in language that reflects the limits of the test.

This is the approach adopted by the Centre for Forensic Neuroscience. Every examination is assessed for suitability, structured according to APA professional standards, and reported with an honest acknowledgement of the test’s strengths and limitations.

Conclusion

Polygraph accuracy figures from meta-analytic research are meaningful and important. They show that well-validated polygraph techniques can discriminate between deceptive and truthful responses with a degree of reliability that exceeds chance and exceeds unaided human judgement.

But those figures do not, by themselves, tell you the probability that a particular person is lying or telling the truth. That probability depends on the base rate — the prevalence of the behaviour being tested for in the relevant population. When the base rate is high, positive results carry strong predictive value. When the base rate is low, false positive rates increase, and positive results should be interpreted with greater caution.

A scientifically responsible approach to polygraph testing acknowledges these realities. It does not dismiss polygraph examination as unreliable, and it does not present it as infallible. It treats each result as a probabilistic contribution to a wider evidence base — a contribution that is most valuable when the test is properly structured, professionally conducted, and interpreted in the context of the case.

This article is provided for general information and educational purposes. A polygraph examination is not a clinical diagnosis, legal verdict, or guarantee of truthfulness. Results should be interpreted cautiously and in the context of each case. Where safeguarding, legal, or mental health concerns are involved, appropriate professional advice should always be sought.

Dr Keith Ashcroft is a Chartered Psychologist, polygraph examiner, and member of the American Polygraph Association at the Centre for Forensic Neuroscience. If you would like to discuss whether a polygraph examination is appropriate for your circumstances, or if you require advice on interpreting polygraph evidence, please contact Dr Ashcroft for a confidential consultation.

Frequently Asked Questions