A few can catch a liar

Research suggests that most people cannot tell from demeanor when others are lying. Such poor performance is typical not only of laypeople but also of most professionals concerned with lying. In this study, three professional groups with special interest or skill in deception, two law-enforcement groups and a select group of clinical psychologists, obtained high accuracy in judging videotapes of people who were lying or telling the truth about their opinions. These findings strengthen earlier evidence that some professional lie catchers are highly accurate, and that behavioral clues to lying are detectable in real time. This study also provides the first evidence that some psychologists can achieve high accuracy in catching lies.

Participants viewed 10 brief videotaped segments of individuals lying or telling the truth about their opinions and made veracity judgments within 10 seconds. The videotapes contained known behavioral cues to deception including facial expressions, voice, and body language from high-stakes situations.

Background & Context

This 1999 study by Paul Ekman and colleagues represents a pivotal moment in deception detection research, challenging the longstanding finding that most people—including professionals—perform near chance levels when judging truth from lies. Prior research consistently showed that most people cannot tell from demeanor when others are lying, with poor performance typical not only of laypeople but also of most professionals concerned with lying. The study built upon Ekman and O'Sullivan's 1991 work showing videotapes to over 330 law-enforcement personnel from seven different groups, finding that only the U.S. Secret Service was quite accurate.

The central question was whether the Secret Service finding was an anomaly or whether other specialized professional groups could also achieve high accuracy. This research emerged from Ekman's broader "Wizards Project" (originally the Diogenes Project), which sought to identify individuals with exceptional lie detection abilities from demeanor cues. The stakes were high: if behavioral cues to deception could be reliably detected in real time by certain professionals, it would have profound implications for law enforcement, national security, and psychological practice.

Research Design & Methodology

Researchers constructed a 10-item test showing 10 different people, half truthful and half lying about their opinions, with both pro and con positions represented regardless of veracity, displayed as face-and-shoulder close-ups with full audio. The videotapes were specifically selected to contain behavioral clues known from prior research to provide a basis for accurate deception judgments. Observers made their judgments during a 10-second interval after viewing each person.

The study tested three professional groups with specialized interests in deception:

• 23 federal law enforcement officers, primarily from the CIA, with the remainder from other federal agencies
• 43 Los Angeles County sheriffs identified by their department as having special skills
• A select group of clinical psychologists with expertise in deception

The videotaped subjects were lying or telling the truth about personally held opinions—a high-stakes scenario where emotional arousal (fear, guilt, excitement) could betray deception through behavioral leakage. This contrasted with low-stakes laboratory studies where emotional cues are minimal.

Results & Key Findings

The federal officers group achieved the highest accuracy, deciding correctly 73% of the time on average—at its 1999 publication, this 73% figure eclipsed all earlier results to become the highest published lie detection accuracy rate in a literature of over 100 studies. Both the sheriffs and clinical psychologists also obtained high accuracy, though specific percentages for these groups varied.

A striking pattern emerged in the data:

• The federal officers detected lies 80% of the time but detected truths only 66% of the time
• The same pattern held true for both groups of clinical psychologists and the Los Angeles County sheriffs
• The most accurate groups were better at detecting lies than detecting truths
• A substantial number of people in every category performed at or below chance

These findings strengthened earlier evidence that some professional lie catchers are highly accurate and that behavioral clues to lying are detectable in real time, while also providing the first evidence that some psychologists can achieve high accuracy in catching lies.

Discussion & Significance

This study provided compelling evidence that deception detection from behavioral cues is not uniformly at chance level—certain professionals with specialized training, experience, or motivation can exceed typical performance. Secret Service agents, who are "always looking for a needle in a haystack in a crowd," may have increased alertness that enhances their detection abilities. The finding that clinical psychologists could also achieve high accuracy was particularly novel, suggesting that expertise in reading emotions and nonverbal behavior translates to lie detection skill.

The asymmetry between lie and truth detection—where the best performers were more accurate at spotting deception than veracity—suggests these groups may have developed a heightened sensitivity to behavioral cues indicative of deception, possibly at the cost of a slight bias toward suspicion. This pattern aligns with the occupational demands of law enforcement and investigative work, where missing a lie carries greater consequences than false suspicion.

The research contributed to the broader understanding that lie detection is not simply about general "ability" but involves specific skills, training, motivation, and perhaps innate talent. Within Ekman's larger Wizards Project, only 50 people out of 20,000 tested had the ability to spot deception without formal training, underscoring the rarity of exceptional natural ability while also demonstrating its existence.

Limitations & Considerations

The study has faced methodological critiques. Psychologists Charles F. Bond and Ahmet Uysal criticized the methodology and suspected the performance of reported "wizards" might be due to chance (a type I error). Bond later revealed that the 23 federal officers took three lie detection tests, not one—they achieved 73% accuracy on opinions, 63.4% on a mock theft, and only 48.2% on emotions, with the theft and emotion tests not mentioned in the 1999 article.

Additional concerns include sample size (particularly the 23 federal officers), the use of opinion lies rather than transgression lies typical in criminal contexts, and questions about the ecological validity of brief video clips versus real-world interrogations. Critics noted that some respondents scored their own detection tests, or achieved high scores on some tests but not others, and that finding 29 "wizards" in a cumulated sample of 12,000 participants can be explained by chance alone. The videotaped stimuli, while containing known behavioral cues, may not represent the full complexity of deception in natural settings.

Practical Applications

This research validated the premise that specialized training and professional experience in certain high-stakes environments can enhance lie detection ability. Following publication, Ekman was frequently asked to train law enforcement officers in lie detection, with the biggest requests coming from the Bureau of Alcohol, Tobacco and Firearms. The findings supported development of training programs focused on recognizing microexpressions and behavioral leakage cues.

For consumers and professionals, the study offers a nuanced message: while exceptional lie detectors exist, they are rare, and even trained professionals show considerable individual variation. Even judgments by professionals with training in deception will not always be accurate, and without proper training and motivation, most people do quite poorly—judging deception from facial expressions and body language will probably never be sufficiently accurate to be admissible in the courtroom. The asymmetry between lie and truth detection also cautions against over-reliance on behavioral cues alone, as heightened lie detection may come with increased false suspicion of truthful individuals.