Hypnotically induced belief in lies: impacts on microexpressions during deception and implications for lie detection

This study aimed to investigate whether microexpressions differ between lying (when believing the lie is true) and truth-telling, and to explore the feasibility of lie detection using microexpression changes. A 2×3 within-subjects design was adopted, with 30 college students screened for high hypnotic susceptibility via the Harvard Group Scale of Hypnotic Susceptibility (HGSHS-A; scores 9–12). Participants told real and hypnotically induced false stories across three states: waking state 1, hypnotic state (with belief in the lie), and waking state 2. Facial microexpressions were coded using the Facial Action Coding System (FACS), and data were analyzed via mixed-effects models and robust binary logistic regression. Mixed-effects model results revealed significant main effects of state (F = 13.968, p < 0.005, η2 = 0.32) and story type (F = 10.856, p < 0.005, η2 = 0.28), as well as a significant state×story type interaction (p < 0.005), indicating that microexpression differences between lying and truth-telling were most pronounced in the hypnotic state. However, robust binary logistic regression showed that microexpression variables (e.g., action unit frequency, eye movements, blink rate) failed to predict deception (χ2 = 6.30, p = 0.643). To validate the hypnotic state, a tripartite verification (behavioral confirmation, post-hoc self-report, and result consistency) was conducted, confirming effective induction of belief in lies. These findings suggest that hypnotically induced belief in lies modulates microexpression patterns during deception, but microexpressions alone are insufficient for reliable lie detection. This study provides empirical insights into the interplay between belief and deceptive communication in high hypnotic susceptibility populations, and offers practical implications for lie detection in real-world contexts.