Training police officers to detect deceptive eyewitness statements. Does it work?

Köhnken (1987) examined whether experienced police officers could discriminate between truthful and deceptive eyewitness statements and whether training could enhance detection accuracy. The study found that officers performed at below-chance levels (47% accuracy) and that training did not improve performance compared to controls.

Eighty male police officers with 3–26 years of experience were randomly assigned to training versus control conditions and viewed videotaped eyewitness statements (truthful and deceptive) to make veracity judgments. Training focused on behavioral and nonverbal cues to deception.

Köhnken's 1987 study investigated the assumption that experienced police officers possess superior lie-detection skills and whether training could enhance those abilities. Using videotaped eyewitness statements in a realistic investigative context, 80 male officers with varying tenure were randomly assigned to receive training on behavioral deception cues or serve as controls. Results revealed officers performed at below-chance accuracy (47% overall), with better detection of truthful statements (63%) than deceptive ones (36%), and training produced minimal to no improvement. Notably, years of professional experience correlated negatively with accuracy, while age showed positive correlation independent of experience, suggesting field experience may reinforce counterproductive judgment biases rather than improving deception detection.

The study challenges law enforcement assumptions about experiential expertise in deception detection and raises serious questions about the effectiveness of behavioral training interventions for improving witness credibility assessment. Findings suggest that routine reliance on officer judgment for credibility determinations may introduce systematic bias and that alternative, more structured assessment methods warrant consideration in investigative practice.

Background & Context

During the mid-1980s, law enforcement agencies increasingly sought scientific methods to improve officers' ability to detect deception during investigations. Günter Köhnken's 1987 study emerged during this period, examining both whether police officers with field experience could discriminate between truthful and deceptive eyewitness statements and whether training could enhance this ability. This research represented an important early empirical test of assumptions that experienced officers possessed superior lie detection skills and that those skills could be improved through training.

The study addressed a critical gap in forensic psychology: law enforcement professionals regularly make consequential judgments about witness credibility, yet little research had examined their actual accuracy rates or the effectiveness of training interventions. Unlike previous deception studies that focused on attitudes or opinions about private matters, Köhnken designed his study to more closely mirror actual investigative contexts by using factual statements about witnessed events.

Research Design & Methodology

The study examined 80 male police officers with between 3 and 26 years of field experience. Officers were divided into experimental and control groups to test the effectiveness of training on deception detection accuracy.

The methodology employed several important modifications to standard lie-detection paradigms: it used factual statements about events rather than private feelings or attitudes, presented longer and complete statements instead of brief 20-30 second message segments, and required only one judgment per sender rather than multiple within-sender judgments. Participants viewed videotaped eyewitness statements—some truthful, some deceptive—and made veracity judgments about each statement.

Key design elements included:

• Random assignment to training versus no-training control conditions
• Videotaped stimulus materials of actual eyewitness-type statements
• Training focused on behavioral and nonverbal cues to deception
• Assessment of demographic variables including age and years of experience

Results & Key Findings

The central finding was striking: police officers did not perform better than chance expectations in detecting deception, and critically, training did not increase accuracy levels compared to the no-training control group. When officers watched videotaped statements of witnesses, the mean overall accuracy was .47 (with .50 being chance level), consisting of .36 accuracy for false statements and .63 for truthful statements.

Performance below chance: Officers' overall accuracy of 47% fell below the 50% expected by random guessing, indicating systematic biases in their judgments.

Additional key findings:

• Lie-detection accuracy was positively correlated with age but negatively associated with length of professional experience
• Officers showed better accuracy identifying truthful statements (63%) than deceptive ones (36%)
• The training produced minimal to no increase in accuracy, aligning with other early studies showing limited effectiveness of deception detection training

Discussion & Significance

Köhnken's findings challenged the prevailing assumption that police officers develop superior deception detection skills through professional experience. The negative correlation between years of service and accuracy suggested that field experience might actually reinforce counterproductive strategies or biases rather than improving judgment. The positive correlation with age—independent of experience—hinted that general life maturity might be more beneficial than occupational training.

The study's failure to find training effects was particularly significant for law enforcement policy. It suggested that simply educating officers about behavioral cues to deception was insufficient to improve their judgments and might even be counterproductive. This aligned with subsequent research showing that some training programs actually reduced accuracy or increased overconfidence. The results implied that officers' poor performance stemmed not from lack of information but from reliance on invalid cues or systematic biases in processing deception-relevant information.

This research became a foundational citation in the deception detection literature, frequently referenced in discussions about the limitations of traditional behavioral cue training and the need for alternative approaches focused on strategic interviewing and verbal content analysis rather than nonverbal behavior observation.

Limitations & Considerations

The study used videotaped statements rather than live interactions, which may have eliminated important contextual and interactive cues available in actual investigations. The focus on eyewitness statements, while ecologically relevant, represented only one type of deception scenario officers encounter. Additionally, the specific content and nature of the training intervention—while focused on behavioral and nonverbal cues—was not extensively detailed in available abstracts, making it difficult to assess what specific training approach was tested.

The all-male sample and the 1987 publication date mean the findings may not fully generalize to contemporary, more diverse police forces. The videotape technology and quality of that era also differed substantially from modern high-definition video used in current research. Finally, the study examined detection of statements about witnessed events rather than denials of personal involvement in crimes, which present different cognitive and emotional dynamics.

Practical Applications

Köhnken's findings have important implications for law enforcement training and practice. They suggest that traditional approaches emphasizing observation of behavioral "tells" are unlikely to improve officer performance and may foster false confidence. Instead, training should focus on evidence-based interviewing techniques, strategic use of information, and analysis of statement content rather than demeanor. Officers should be made aware that field experience alone does not guarantee expertise in deception detection and may reinforce invalid beliefs.

For police administrators and policymakers, these results underscore the need for careful evaluation of training programs before widespread implementation. Deception detection training should be grounded in empirically validated methods and regularly assessed for effectiveness. The study also highlights the importance of corroborative evidence and investigative procedures that do not rely solely on subjective credibility judgments, which are demonstrably unreliable even among experienced professionals.