Detecting true and false opinions: The Devil's advocate approach as a lie detection aid

This 2010 study by Leal, Vrij, Mann, and Fisher introduced and empirically tested the Devil's Advocate approach — a structured two-question interview technique designed to detect lies about personal opinions. By exploiting the cognitive accessibility asymmetry between supporting and opposing one's true views, the method successfully differentiated truth tellers from liars across both behavioral measures and qualitative observer judgments.

Two-experiment laboratory study using 17 truth tellers and liars in a structured Devil's Advocate interview (Experiment 1, measuring word count and response latency) and 25 independent observers rating qualitative features of the recorded interviews (Experiment 2), with ANOVA used for statistical analysis.

This laboratory study introduced the Devil's Advocate approach to detect deception about personal opinions, addressing an underexplored area of lie detection research. The cognitive-based technique exploits confirmation bias by having participants argue both for and against their stated positions. Two experiments measured behavioral responses and human observer judgments. Results supported the hypothesis that truth-tellers show greater fluency when supporting genuine beliefs, while liars reveal their actual opinions when forced to argue the counter-position. The findings demonstrate a novel application of cognitive load principles to opinion-based deception detection.

This research expands deception detection beyond event-based lies to opinion-based deception, potentially valuable for investigations involving ideological sympathies or belief systems. The cognitive approach offers an alternative to physiological polygraph methods for detecting deception about personal beliefs and attitudes.

Background & Context

For decades, deception researchers have focused primarily on detecting lies about events — whether a suspect was present at a crime scene, whether they performed a specific action. Yet a significant and underexplored category of deception involves lying about one's opinions: feigning support for a cause, concealing ideological sympathies, or misrepresenting personal beliefs. This study by Leal, Vrij, Mann, and Ronald P. Fisher directly targets this gap, introducing and evaluating an interview-based technique called the Devil's Advocate approach.

The research sits within the broader cognitive approach to lie detection, a paradigm championed particularly by Aldert Vrij and colleagues at the University of Portsmouth. Rather than relying on physiological arousal signals — as traditional polygraph techniques do — the cognitive approach exploits the greater mental effort required to fabricate convincing accounts. Vrij argues that the key to deception detection is cognitive load, noting that lying involves a series of cognitive operations including recalling a story without rich mental associations, inhibiting signs of anxiety, and monitoring the target to ensure the false account is being accepted.

Lying about opinions presents a unique cognitive challenge. Unlike lying about events (where a liar must construct a plausible false narrative), people normally think more about reasons that support rather than oppose their opinion, so truth tellers are expected to provide more information and shorter latency times in their responses to the opinion-eliciting question than to the Devil's Advocate question — whereas liars are expected to reveal the opposite pattern, as the Devil's Advocate question is more compatible with their actual beliefs. This asymmetry forms the theoretical foundation of the entire technique.

Research Design & Methodology

The study, authored by Sharon Leal, Aldert Vrij, Samantha Mann, and Ronald P. Fisher — all based at the University of Portsmouth — examined the efficacy of this new approach to detect truths and lies in expressing opinions. The research was structured as two sequential experiments, each targeting a distinct level of analysis: behavioral measurement and qualitative human judgment.

Key design elements of the two experiments included:

• Experiment 1: Interviewees were first asked an opinion-eliciting question that asked participants to argue in favour of their personal view, followed by a Devil's Advocate question that asked participants to argue against their personal view.
• Sample (Exp. 1): Seventeen participants assigned to truth-telling or lying conditions, interviewed individually.
• Behavioral measures: Number of words produced and latency times (response onset delays) to each question type were recorded and compared between conditions.
• Experiment 2: Twenty-five independent observers viewed the recorded interviews and made qualitative judgments about immediacy, plausibility, and emotional involvement in each response.
• Statistical approach: Analysis of Variance (ANOVA) was used to assess differences across conditions and question types.

Participants were adult men and women, and the study was conducted under controlled laboratory conditions. Critically, the design exploited a well-established psychological principle — confirmation bias in opinion formation — to create measurable, predictable asymmetries between honest and deceptive responders when confronted with counter-attitudinal questioning.

Results & Key Findings

The core hypotheses were supported in both experiments. Truth tellers and liars displayed systematically different behavioral and qualitative profiles in response to the two question types, validating the underlying theoretical model of opinion-based deception.

From Experiment 1 (behavioral data):

• Truth tellers produced more words and had shorter response latencies when answering the opinion-eliciting question (arguing for their real view) than when answering the Devil's Advocate question.
• Liars showed the reverse pattern — they were more fluent and faster when answering the Devil's Advocate question (which actually aligned with their true beliefs) than the opinion-eliciting question.
• This crossover interaction confirms that the technique captures genuine differences in cognitive accessibility of supporting versus opposing arguments.

From Experiment 2 (observer judgments):

• Truth tellers' opinion-eliciting answers were seen as more immediate, plausible, and revealed more emotional involvement than their Devil's Advocate answers.
• No clear qualitative differences emerged in liars' answers to the two question types — consistent with the hypothesis that liars would find both questions similarly manageable (one matching their true view, one requiring fabrication).
• The pattern of observer ratings cleanly distinguished truth tellers from liars at the level of qualitative impression, not just objective word counts.

Overall, truth tellers provided more information and shorter latency times in their responses to the opinion-eliciting question than to the Devil's Advocate question, while liars' responses revealed the opposite pattern — as the Devil's Advocate question was more compatible with their actual beliefs. The study thus demonstrated that a two-question structural interview format can reliably surface deception about opinions through both quantitative and qualitative channels.

Discussion & Significance

This research makes a meaningful contribution by extending cognitive lie detection into a domain — opinion expression — that had been largely unaddressed. Most prior deception research focused on event-based lies, where suspects fabricate accounts of their whereabouts or actions. Research on lies about opinions is scarce, with the Devil's Advocate approach being a noticeable exception, and the approach had been examined only twice with single interviewees at the time of earlier follow-up studies, underscoring the need for replication. Leal et al.'s 2010 paper was the pioneering work that opened this line of inquiry.

The significance of the approach lies in its structural elegance. Unlike techniques that require specialized equipment, physiological monitoring, or expert analysis of micro-expressions, the Devil's Advocate method can be implemented with nothing more than a structured interview protocol. This makes it accessible and scalable across diverse investigative contexts — from security screening to credibility assessments in legal and employment settings.

The study also contributes to a growing body of evidence that strategically designed questioning — rather than passive observation of behavioral cues — is a more productive route to deception detection. Vrij, Granhag, Mann, and Leal later summarized this trajectory in a paper titled "Outsmarting the Liars: Toward a Cognitive Lie Detection Approach," published in Current Directions in Psychological Science. The 2010 study by Leal et al. was a foundational empirical pillar of that emerging framework.

Limitations & Considerations

As a pioneering laboratory study, this research carries several important caveats that should temper its direct application:

• Small sample sizes: Experiment 1 involved only 17 participants, and Experiment 2 used 25 observers — numbers too small to draw robust conclusions about effect stability or generalizability.
• Controlled laboratory setting: Participants in analog studies face considerably lower stakes than real-world suspects; motivational differences may affect how naturally the predicted behavioral patterns emerge.
• Opinion topic specificity: The technique depends on participants holding genuinely strong opinions about the topic under inquiry — its efficacy may diminish when topics are less personally salient.
• Observer judgment subjectivity: Experiment 2 relied on lay observers' qualitative ratings, which introduces potential for variability in how different raters perceive immediacy, plausibility, and emotional engagement.
• Replication need: The Devil's Advocate approach has been examined only twice with single interviewees and needs replication across diverse samples, topics, and cultural contexts.

The authors themselves note the technique is promising but preliminary, explicitly calling for future research to consolidate and extend these findings. Subsequent work by Deeb et al. (2018) began addressing the paired-suspect consistency dimension, and more recent studies have explored combining the Devil's Advocate approach with the Verifiability Approach to strengthen detection power.

Practical Applications

For investigators and interviewers working in security, intelligence, or law enforcement contexts, this research offers a concrete, low-cost interview strategy for situations where a subject's true ideological or attitudinal stance is in question. The protocol is straightforward: first ask the interviewee to argue for the position they claim to hold, then ask them to argue against it. The relative fluency, word count, latency, and qualitative richness of the two answers provides diagnostic information without the need for instrumentation. Truth tellers are expected to be more eloquent, passionate, and less predictable than lie tellers in responding to the opinion-eliciting question, and the difference in these qualities between the two question types is expected to be more pronounced in truth tellers than in lie tellers.

For polygraph practitioners and credibility assessment professionals, the Devil's Advocate approach represents a complementary verbal technique that can be integrated into pre-test or post-test interview phases — particularly when cases involve suspected extremist sympathies, undisclosed affiliations, or feigned attitudes. While not a replacement for validated psychophysiological methods, it exemplifies how structured interview design can independently yield meaningful deception-detection signal from verbal behavior alone.