Police Officers’ Lie Detection Accuracy: Interrogating Freely Versus Observing Video

This 2004 study examined whether experienced police officers' lie detection accuracy differs when actively interrogating suspects versus passively observing videotaped interrogations. Thirty officers either conducted free-style interrogations of mock guilty or innocent suspects or watched video recordings of interrogations, then made veracity judgments.

Between-subjects experimental design with 30 experienced police officers who either conducted live interrogations of college student mock suspects (guilty or innocent of a simulated crime) or observed videotaped versions of interrogations. Officers used their preferred interrogation techniques and made post-interrogation veracity judgments.

Background & Context

This 2004 study addresses a fundamental question in deception detection research: does actively interrogating a suspect improve lie detection accuracy compared to passively observing an interrogation? The research investigated experienced police officers' lie detection accuracy when conducting interrogations of suspects versus observing videotaped interrogations. Understanding this distinction is critical because real-world investigations depend heavily on officers' ability to accurately assess credibility during direct interactions with suspects.

Prior research had revealed a puzzling pattern: less involved lie catchers (observers) are often at least as accurate and sometimes even more accurate in detecting lies than highly involved lie catchers (interrogators). This phenomenon, known as the "interviewer paradox," suggested that the very act of conducting an interrogation might impair judgment. The study fills an important gap by examining whether police officers—who possess professional training and experience—demonstrate the same pattern observed in laboratory studies with civilians.

Research Design & Methodology

The study involved 30 experienced police officers who were assigned to one of two conditions. Each police officer either conducted an interrogation of a college student acting as a suspect who was either guilty or innocent of a mock crime, or observed videotaped versions of interrogations. The suspects had either committed a mock theft or completed an innocent alternative activity.

The police officers conducting live interrogations had the opportunity to conduct the interrogation in the manner of their own choice, allowing researchers to observe naturalistic interviewing behaviors. Key features of the design included:

• 30 experienced police officers as judges
• College student actors as mock suspects (guilty and innocent)
• Free-style interrogation approach without experimental constraints
• Between-subjects comparison: active interrogators vs. passive video observers
• Post-interrogation veracity judgments and confidence ratings

The interrogators reported to rely on verbal content more when interrogating than when watching video, and researchers analyzed both accuracy rates and the cues officers reported using to make their judgments.

Results & Key Findings

The police officers failed to detect deception better than chance. This striking finding demonstrates that professional experience does not guarantee accurate deception detection under typical interrogation conditions.

There was no difference in accuracy between police officers interrogating live and observing video. The expected advantage of direct interaction did not materialize. Both groups performed at approximately chance levels.

Critical patterns emerged in officer behavior and judgment:

• Chance-level accuracy for both interrogators and video observers
• No advantage for active interrogation over passive observation
• Truth bias: interrogators had a more pronounced truth bias than observers, meaning they tended to judge messages as truthful more often than deceptive
• Differential cue use: interrogators and observers attended to different types of information when evaluating the target—interrogators relied more on nonverbal behavior, whereas observers attended to verbal behavior

Discussion & Significance

The findings challenge the assumption that direct interaction enhances deception detection ability. Despite having control over question selection, pacing, and conversational dynamics, police interrogators achieved no better accuracy than passive observers. This contradicts the intuitive belief that skilled questioning and rapport-building during face-to-face interrogation provide diagnostic advantages.

The pronounced truth bias among interrogators represents a particularly concerning finding for real-world practice. Officers conducting interrogations showed a systematic tendency to believe suspects, which may stem from conversational norms that presume honesty or from interpersonal dynamics that develop during direct interaction. Combined with chance-level accuracy overall, this bias means officers may be especially vulnerable to missing deception when it occurs.

The differential reliance on nonverbal versus verbal cues helps explain the null findings. Interrogators' greater reliance on nonverbal behavior is problematic because research consistently shows that nonverbal cues are weak and unreliable indicators of deception. Observers' focus on verbal content may be equally unproductive without systematic frameworks for evaluating statement consistency and plausibility.

Limitations & Considerations

The mock crime paradigm, while allowing controlled comparison, differs from actual criminal investigations in stake and consequence. College students acting as suspects may not display the same behavioral patterns as genuine criminal suspects facing real consequences. The study examined relatively brief interrogations without access to case evidence, which limits ecological validity.

The sample size of 30 officers, while adequate for detecting large effects, may have limited power to detect smaller differences between conditions. Additionally, officers were allowed to use their preferred interrogation techniques without standardization, which introduces variability but also reflects real-world practice.

Practical Applications

This research has profound implications for criminal justice practice. It demonstrates that neither experience nor direct interaction guarantees accurate credibility assessment. Police departments should recognize that traditional interrogation approaches do not reliably produce accurate deception judgments and that officers' confidence may exceed their actual accuracy.

The findings support the development and adoption of evidence-based interviewing techniques, such as the Strategic Use of Evidence (SUE) approach that emerged from this research program. Rather than relying on intuitive judgments about demeanor and nonverbal behavior, investigators should focus on strategic questioning techniques that elicit verifiable information and create statement-evidence inconsistencies. Later research showed that SUE-trained interviewers achieved 65% accuracy compared to 43% for untrained interviewers, demonstrating that structured, evidence-focused methods can overcome the limitations revealed in this study.