Interview anxiety: what experiments recorded in innocent people
Interview anxiety can make an innocent person hesitate, fidget or look away. Those behaviors reveal discomfort without identifying its cause. In the FBI's Law Enforcement Bulletin in 2014, Brian D. Fitch described speech errors, restless movement and gaze aversion among innocent people under scrutiny. Unfamiliarity with criminal procedure, accusations and an aggressive manner can all increase stress. The visible movement may be real while the inference of guilt is mistaken. The distinction concerns the origin of the stress, not whether the person appears convincingly calm.
Fitch also distinguished gathering information from pressing for an admission. An interview that encourages a detailed narrative produces material that can be compared with other evidence. An accusatory approach can itself generate the fear that an observer then interprets as concealment. His discussion of interview anxiety was professional guidance drawing on research, rather than a new experiment establishing a diagnostic gesture. It did not make a stammer, a restless hand or averted eyes a reliable substitute for checking what happened.
A 2016 PLOS ONE study by Glynis Bogaard and colleagues investigated the beliefs behind those judgments. Its 199 participants included 95 Dutch police officers and 104 undergraduates. The officers were detectives or professional interrogators with an average of 22 years' experience. They and the students first supplied their own ideas about detecting deception, then evaluated 28 features of verbal content. The design measured what people believed, allowing the researchers to compare those beliefs with published evidence about actual differences between truthful and fabricated statements.
Both groups frequently supplied stereotypical signs such as looking away, nervousness and sweating. Their spontaneous suggestions emphasized nonverbal behavior more than spoken content. The researchers also found that the officers recognized some useful verbal features better than students. That narrower advantage did not remove their unsupported beliefs about outward nervousness. Experience in interviewing and the accuracy of a particular belief were separate measurements. The survey exposes one reason interview anxiety becomes incriminating in practice: an observer may bring a confident expectation to behavior that lacks the assumed diagnostic value.
A different primary experiment, published by Sole Yoo and colleagues in PLOS ONE in 2020, separated the act of deception from the pressure of being scrutinized. Participants played a card game in an MRI scanner with financial incentives to lie on some trials. They were told that observers would judge their facial expressions and physiological signals. In fact, a computer determined which trials would attract scrutiny. Of 22 recruited participants, 19 remained in the analysis after exclusions for performance and movement. The researchers could compare truthful and deceptive trials with and without that pressure.
The team reported anxiety-related brain responses under scrutiny in both truthful and deceptive conditions, while also identifying differences between those conditions. That finding supports the distinction between anxiety and deception; it does not establish a visible test of innocence. The investigators acknowledged the small sample and the absence of anxiety ratings for individual trials. Their laboratory procedure was also a card game, rather than an actual criminal investigation. Interview anxiety involves processes that can accompany both honesty and concealment, and a physiological response cannot simply be equated with the deceptive act.
A 2024 Frontiers in Psychology experiment examined behavior in a staged theft involving 50 Chinese participants, divided evenly between truthful and deceptive conditions. Truthful participants took part in a poker-room scene; deceptive participants took money from a wallet and were supplied with a description of the scene to imitate. Both groups then faced an interviewer wearing a police uniform who did not know their assigned condition. The researchers recorded the interviews and coded movements, providing a direct behavioral record rather than relying only on observers' beliefs about nervousness.
The deceptive group reported stronger emotions and, with gender and socioeconomic status included as covariates, made more self-directed touching movements. Yet the gaze-duration difference had a p-value of 0.057, falling short of the conventional 0.05 threshold; other coded movements did not significantly separate the groups. The touching result also weakened when the covariates were omitted. This was a limited experimental pattern, not a demonstration that any particular nervous participant was deceptive. The details resist the sweeping claim that interview anxiety always looks the same or that a cluster of familiar gestures settles guilt.
In their 2020 research synthesis, Aldert Vrij and Ronald P. Fisher separated feeling nervous from displaying nervous behavior. Their review found that liars can report greater nervousness without appearing more nervous to observers. They proposed deliberate control and the mental demands of lying as explanations for reduced movement. Those are interpretations of the research, not direct access to another person's motives. The distinction helps explain why a calm presentation can coexist with deception while an anxious presentation accompanies truth.
The same synthesis discusses accusation as a source of nervous behavior in both groups. An interviewer can create pressure and then misattribute the discomfort to dishonesty. It also reports a meta-analysis in which observers classified truth and lies with 52% accuracy from silent video, 63% from audio alone and 56% with both. Those pooled results concern the experimental material in that analysis, rather than every individual or setting. Seeing a person's manner alongside their speech did not automatically produce greater accuracy.
- Bogaard, Meijer, Vrij and Merckelbach, Strong, but Wrong: Lay People's and Police Officers' Beliefs about Verbal and Nonverbal Cues to Deception, PLOS ONE, 2016 — journals.plos.org
- Yoo and colleagues, Neural correlates of anxiety under interrogation in guilt or innocence contexts, PLOS ONE, 2020 — journals.plos.org
- Li, Song, Li and Li, Nonverbal cues to deception: insights from a mock crime scenario in a Chinese sample, Frontiers in Psychology, 2024 — www.frontiersin.org
- Vrij and Fisher, Unraveling the Misconception About Deception and Nervous Behavior, Frontiers in Psychology, 2020, research synthesis — www.frontiersin.org
- Brian D. Fitch, The Truth About Lying: What Investigators Need to Know, FBI Law Enforcement Bulletin, 10 June 2014 — leb.fbi.gov