Researchers studying face-to-face conversations found that higher psychopathic traits correlated with reduced affective sharing and, to a lesser degree, reduced physiological synchrony with a conversation partner but not with lower accuracy in inferring the other person’s emotional state.
Most empathy research asks participants to view photographs or pre-recorded video clips of strangers. Psychopathy research relies on this approach almost exclusively, using self-report questionnaires or image-based tasks rather than actual social interaction. Researchers at Victoria University of Wellington argued this misses something fundamental: empathy unfolds between people in real time. Studying it in isolation may not reflect how it works in genuine social contexts.
Prior work mainly from questionnaire studies and meta-analyses established that psychopathy associates with deficits in both affective empathy (sharing others’ feelings) and cognitive empathy (accurately inferring others’ emotional states), with affective deficits typically larger. Separately, studies on physiological synchrony — the tendency for two people’s autonomic responses to align during interaction found it correlates with empathy in couples, therapist-patient pairs, and mother-child interactions. No prior study had examined physiological synchrony in relation to psychopathic traits.
Researchers recruited 53 community-dwelling dyads from the Wellington, New Zealand area through flyers, social media, and a university participant pool. Technical problems during data collection eliminated 12 dyads, leaving 41 dyads (82 individuals) in the final analyses. Individual outcome analyses used between 33 and 36 dyads depending on data availability. About half the pairs knew each other beforehand as friends or romantic partners; the rest were strangers. Participants ranged in age from 18 to 78, and about 71 percent were women.
Each participant wore a wireless physiological recording vest that captured heart rate and electrodermal activity (EDA, a measure of skin conductance related to arousal). Each pair then held four six-minute conversations, prompted by a written topic: a major positive life event, a major negative life event, a major regret, and a moment of pride. Researchers video-recorded all conversations.
Afterward, each participant watched the recordings twice — once rating their own emotional intensity moment by moment, and once rating their partner’s. The researchers computed empathic accuracy by correlating a person’s self-ratings with their partner’s ratings of them. They measured affective sharing by correlating the two individuals’ self-ratings with each other. For physiological synchrony, they used a statistical method that accounts for autocorrelation and timing lags in the data. Each participant also completed the PPI-R-40, a validated self-report measure of psychopathic traits across three factors: fearless dominance, self-centered impulsivity, and coldheartedness.
The study was not pre-registered, making all analyses exploratory. To guard against selective reporting, the researchers tested each finding across four different analytic methods simultaneously.
Participants showed meaningful empathic accuracy on average, including between strangers. Familiarity significantly predicted higher empathic accuracy across all four analytic approaches. Psychopathic traits showed no significant relationship with empathic accuracy in any of them.
For affective sharing, familiarity had no significant effect — people showed emotional resonance with their partner regardless of whether they knew each other. The self-centered impulsivity factor of psychopathy was negatively associated with affective sharing, and this held across all four analytic methods.
For physiological synchrony, heart rate showed no evidence of synchronization between partners, ruling out further analysis of that measure. EDA did synchronize significantly on average. The coldheartedness factor of psychopathy negatively associated with EDA synchrony, but only reached statistical significance in two of the four analytic approaches. The researchers note this pattern “should consequently be interpreted with caution.”
The authors argue their findings align with the longstanding view that psychopathy involves affective but not cognitive empathy deficits. They caution, though, that the null finding on empathic accuracy may reflect the richer contextual information available in live interaction. They write that “in real-world situations, individuals higher in psychopathic traits might exhibit fewer or no differences in empathic accuracy because richer contextual cues can facilitate accurate inferences about others’ emotions.” They also note this is, to their knowledge, the first study to examine physiological synchrony in relation to psychopathic traits.
The researchers are explicit that these findings are preliminary. The sample of 41 dyads limits the ability to detect small effects. The study did not control for gender composition of pairs, nor did it record how long familiar pairs had known each other. Conversations varied substantially in dynamics some participants dominated more than others because the naturalistic design limits control over what people actually say. The study was not pre-registered. The researchers explicitly call for replication before drawing strong conclusions.
References
Matthias Burghart, Roydon Goldsack, Areito Echevarria, and Hedwig Eisenbarth, “Empathy, physiological synchrony, and psychopathy: preliminary insights from naturalistic dyadic interactions,” Cognition & Emotion, published online March 2, 2026. https://doi.org/10.1080/02699931.2026.2637546
