Deception Detection: Interrogators' and Observers' Decoding of Consecutive Statements

This laboratory study examined which cues evaluators use when assessing deception across repeated interrogations and whether face-to-face interrogators differ from video observers. Twenty-four suspects (12 liars, 12 truth-tellers) were interrogated three times over 11 days, with veracity assessed by interrogators and observers. Results showed consistency over time was the most commonly used cue, but its predictive accuracy was alarmingly low.

Laboratory analog design with 24 participants witnessing a staged event, then interrogated three times (at 3 hours, 4 days, and 11 days). Half lied about the event while half told the truth. Veracity judgments were made by 24 interrogators and 144 video observers, with justifications coded for cue usage.

Background & Context

This 2001 study by Granhag and Strömwall addresses two forensically important but previously neglected issues in interpersonal deception: which cues lie catchers use when accessing repeated interrogations, and how face-to-face interrogators differ from observers. The research emerged from a critical gap in deception detection literature that had predominantly focused on single interrogation encounters, despite the reality that suspects in criminal investigations typically face multiple interviews.

The work is particularly significant because it empirically tests what Granhag and Strömwall termed the "consistency heuristic"—the widely-held assumption that inconsistency implies deception and consistency implies truth. This belief pervades not only lay intuitions but also professional interrogation manuals and legal practice. The researchers sought to determine whether this heuristic actually has diagnostic value in distinguishing truth-tellers from liars across repeated interrogations.

Research Design & Methodology

After watching a staged event, 24 suspects (12 liars and 12 truth tellers) were interrogated three times over a period of 11 days. After the final interrogation, the veracity of each suspect was assessed by his or her interrogator and by 6 observers who had watched the interrogations on video only. The timeline involved interrogations conducted at 3 hours, 4 days, and 11 days after the staged event.

The experimental manipulation was straightforward but ecologically relevant. Half of the witnesses were instructed to be truthful when describing the "crime" and half were asked to lie such that the victim was the perpetrator. This design allowed liars to construct a plausible but false narrative while truth-tellers recounted their genuine observations.

The study employed a dual-perspective methodology, comparing 24 interrogators (each conducting one interrogation series) with 144 observers (six per suspect) who watched videotaped versions of the same interrogations. Participants then made veracity judgments and provided written justifications for their decisions, allowing researchers to code which cues influenced their judgments.

Results & Key Findings

Consistency over time was by far the most commonly used cue for justifying veracity judgments. Critically, the predictive accuracy for this cue was alarmingly low. This headline finding directly challenges one of the most fundamental assumptions in both folk psychology and professional interrogation practice.

Key empirical outcomes included:

• Deceptive statements were at least equally consistent as truthful statements, contradicting the consistency heuristic
• Interrogators used verbal cues to a significantly greater extent than did the observers, reversing patterns from prior research using scripted interviews
• A probing effect was demonstrated—probed suspects were perceived as significantly more honest than nonprobed suspects
• Limited support for a previously reported honesty effect (interrogators perceiving suspects as more honest than observers) was obtained

The researchers found that people tend to disagree as to whether a particular set of consecutive statements is consistent or not, suggesting that consistency itself is a subjective and unreliable judgment criterion. Moreover, both interrogators and observers performed poorly at distinguishing truth-tellers from liars overall.

Discussion & Significance

To explain why liars can be as consistent as truth-tellers, Granhag and Strömwall proposed the "repeat vs. reconstruct" hypothesis, assuming that liars try to repeat their initial statement, whereas truth-tellers try to reconstruct a previously experienced event. This mechanism suggests that liars, motivated to avoid contradiction, simply rehearse and repeat their fabricated story, while truth-tellers engage in genuine memory reconstruction—a process naturally subject to variation and omission over time.

The findings have profound implications for investigative practice. The low predictive accuracy found for the consistency heuristic strongly questions beliefs held by professional lie-catchers, opinions expressed in the psycho-legal literature and recommendations found in applied interrogation manuals. The research demonstrates that access to multiple interrogations does not automatically improve deception detection and may actually increase overconfidence.

The reversal regarding which cues interrogators versus observers use is particularly noteworthy. Unlike previous studies using scripted interviews where interrogators focused on nonverbal behavior, in this naturalistic design interrogators used verbal cues to a significantly greater extent, suggesting that when professionals control the questioning, they attend more carefully to statement content.

Limitations & Considerations

The study employed a laboratory analog design with undergraduate participants as both "suspects" and some evaluators, which limits generalizability to high-stakes criminal investigations. The stakes for deception were relatively low compared to real forensic contexts, potentially affecting both liars' motivation and behavioral cues. Additionally, the staged event was relatively simple and recent, unlike complex crimes investigated after longer delays.

The sample size of 24 suspects (12 per condition), while adequate for the experimental design, represents a modest dataset. The study also did not examine whether certain types of inconsistencies (e.g., peripheral versus central details) might have differential diagnostic value, instead treating consistency as a unitary construct.

Practical Applications

This research carries direct implications for investigative interviewing and credibility assessment. Practitioners should recognize that statement consistency across repeated interviews cannot be reliably used to distinguish truthful from deceptive accounts. Training programs that emphasize consistency analysis as a deception detection tool may be promoting ineffective or even counterproductive strategies. Investigators conducting multiple interviews should avoid assuming that consistent accounts indicate truthfulness or that minor inconsistencies signal deception.

The findings also suggest caution regarding the probing effect—repeatedly questioning suspects may paradoxically increase perceptions of honesty rather than revealing deception. For legal contexts, the research challenges the evidentiary weight often given to consistency in witness testimony and suggests that decision-makers should not rely on consistency heuristics when evaluating credibility.