Catching liars: Training mental health and legal professionals to detect high-stakes lies

Shaw, Porter, and ten Brinke (2013) examined whether a full-day, empirically grounded training workshop could improve the deception detection accuracy of 42 legal and mental health professionals evaluating high-stakes emotional lies. The study found that accuracy increased dramatically from a below-chance baseline of 46.4% to 80.9% post-training, offering rare empirical evidence that structured professional development can produce substantial improvements in credibility assessment.

A within-subjects pre–post design was used with 42 legal and mental health professionals, who viewed high-stakes emotional pleader stimuli and made veracity judgments at baseline and again after receiving a full-day empirically based training workshop on deception detection cues.

This study addressed the critical gap between professionals' need to detect deception and their typically chance-level performance by testing whether targeted training could improve detection of high-stakes lies. Forty-two legal and mental health professionals participated in a comprehensive, empirically-based training workshop that taught behavioral, verbal, and facial cues associated with genuine versus feigned emotion. Using real-world footage of high-stakes emotional pleaders, the researchers found that a single day of training dramatically improved detection accuracy from below-chance levels (46.4%) to substantially above-chance performance (80.9%). The training emphasized a multi-channel approach examining multiple behavioral indicators simultaneously, representing a significant departure from previous studies that used low-stakes scenarios and student samples.

This research demonstrates that brief but intensive, empirically-grounded training can dramatically improve professionals' ability to detect high-stakes deception, suggesting significant potential for enhancing credibility assessment in forensic and clinical settings through targeted professional development.

Background & Context

The ability to reliably detect deception is one of the most consequential skills in forensic, legal, and mental health settings — yet it has proven stubbornly elusive. Although the ability to detect deception is critical in many professional contexts, most observers, including professional lie-catchers, are able to identify deceivers at the level of chance only, and almost all studies of deception detection have used low-stakes scenarios in evaluating both deceptive behavior and training effectiveness.

This chronic underperformance raises an urgent question: can targeted training make a meaningful difference? Literature surrounding the accuracy of deception detection has produced inconsistent findings, and although recent research has suggested that high-stakes situations may produce reliable cues to deception, it remained unclear whether knowledge of these cues actually improves lie detection.

Shaw, Porter, and ten Brinke's 2013 study was designed precisely to fill this gap — testing whether an intensive, empirically grounded training intervention could elevate detection accuracy in professionals who routinely must evaluate credibility in real, consequential situations.

Research Design & Methodology

The researchers evaluated the effectiveness of a comprehensive, empirically based full-day training workshop in improving the ability of 42 legal and mental health professionals to detect extremely high-stakes emotional lies. This represented a notable departure from the student samples and trivial deception scenarios that dominate earlier literature.

The stimulus materials used genuine, real-world footage rather than laboratory-constructed scenarios — a critical design choice that ensured ecological validity. Participants' ability to discriminate sincere and insincere pleaders was measured at both baseline and post-training, creating a within-subjects design that allowed direct before-and-after comparison for each individual.

Key design elements of the study included:

• Sample: 42 practicing legal and mental health professionals
• Format: Full-day, comprehensive empirically based training workshop
• Stimulus type: High-stakes emotional pleaders (sincere vs. insincere)
• Measurement: Deception detection accuracy assessed at baseline and post-training
• Design: Within-subjects (repeated measures) pre–post comparison

The training workshop drew on empirically validated methods, synthesizing research on behavioral, verbal, and facial cues associated with genuine versus feigned emotion in high-stakes contexts — a multi-channel approach consistent with Porter and ten Brinke's broader theoretical framework that emphasizes a holistic approach with concurrent attention to multiple channels of a target's behavior, including illustrators, blink and pause rate, speech rate, vague descriptions, repeated details, and emotional 'leakage' in the face.

Results & Key Findings

The headline result was striking and statistically robust: overall accuracy increased significantly from M = 46.4% to 80.9% — a gain of more than 34 percentage points following a single day of structured training. This moved participants from below-chance performance into genuinely above-chance detection capability.

Key statistical outcomes from the study:

• Baseline accuracy: 46.4% — below the theoretical 50% chance level, consistent with a truth bias in untrained observers
• Post-training accuracy: 80.9% — a substantial and significant improvement
• Absolute gain: ~34.5 percentage points after a single full-day workshop
• Participants shifted from performing worse than chance to performing well above chance
• Training focused on high-stakes emotional lies — a more demanding and ecologically valid target than typical lab studies

The authors cautiously suggest that training professionals to apply empirically validated methods to deception detection can increase their ability to correctly discriminate between liars and truth-tellers. The magnitude of the improvement is especially notable given that the stimulus materials involved genuinely high-stakes emotional deception, which is typically considered harder to decode than low-stakes laboratory fabrications.

Discussion & Significance

The finding that professionals began below chance — at just 46.4% — powerfully replicates and extends prior research demonstrating that expertise in law or mental health confers no natural advantage in lie detection. Several studies have found that parole boards, psychologists, mental health practitioners, and legal professionals detected lies at less than or at chance levels, and this study confirmed that pattern as a baseline. The real contribution is demonstrating that this deficit is correctable.

The post-training accuracy of 80.9% represents a landmark result in a field where even single-digit improvements are often considered meaningful. Comparable research on group-based lie detection has similarly found that structured processes can enhance judgment accuracy, with groups of individuals sometimes making more accurate judgments than the average individual could make alone, though lie detection remains an exceptionally challenging task.

The study's use of real-world high-stakes pleaders — rather than role-playing undergraduates — aligns with growing consensus that training effectiveness must be measured against authentic forensic stimuli. Most prior research was conducted as laboratory experiments where liars had little to lose if caught; if valid cues are typical only to high-stakes situations, it is plausible that instruction to focus on these cues would increase accuracy. Shaw et al.'s results support exactly this hypothesis.

Limitations & Considerations

Several methodological considerations temper the enthusiasm these results deserve. The sample of 42 professionals, while appropriate for an initial proof-of-concept study, is modest, and the generalizability of findings to other professional populations, training formats, or cultural contexts cannot be assumed. The within-subjects design also means that order effects, re-exposure to stimulus materials, and practice-related familiarity could partially inflate the post-training accuracy estimates.

• Sample size: N = 42 is relatively small for strong generalizations
• Single-session training: Long-term retention of training benefits was not assessed
• Stimulus generalizability: Findings centered on a specific type of high-stakes pleader footage
• Within-subjects design: Potential for carryover or practice effects inflating post-test scores
• No active control group: A matched group receiving no training or a placebo intervention would strengthen causal inference

The authors themselves acknowledge these constraints, cautiously suggesting rather than asserting definitive conclusions — a commendable level of scientific restraint given the magnitude of the effect observed. Future research replicating these findings with larger samples, active control conditions, and longitudinal follow-up is essential.

Practical Applications

For forensic practitioners, legal professionals, and mental health evaluators, this research delivers an important and actionable message: training works. The shift from sub-chance to ~81% accuracy through a structured one-day intervention suggests that agencies responsible for credibility assessment — parole boards, forensic interviewers, asylum adjudicators, and courtroom evaluators — could meaningfully improve outcomes by investing in empirically grounded professional development programs focused on high-stakes behavioral cues.

For the broader field of deception detection science, the study reinforces the value of a holistic approach with concurrent attention to multiple channels of a target's behavior, ideally with video for review, and sensitivity to changes from baseline behavior — rather than relying on any single channel or intuitive judgment. Practitioners and consumers seeking reliable credibility assessment methods should prioritize structured, evidence-based training over unvalidated intuition or folk beliefs about lying behavior.