Cross-recurrence quantification analysis captures inter-brain coupling during naturalistic negotiation: a new dynamic approach for hyperscanning.

Naturalistic interactions involve dynamic, nonlinear coordination that unfolds across multiple timescales, yet most hyperscanning studies rely on analytical methods designed for passive, stimulus-locked contexts. We introduce cross-recurrence quantification analysis (CRQA)-a method that treats signals as a coupled dynamical system and characterizes the patterns of their joint trajectories across time-to measure brain-to-brain coupling during a free-flowing negotiation task. Dyads were scanned with fNIRS as they negotiated budget allocations to solve a public health crisis. We measured neural coupling in regions critical for social cognition-medial prefrontal cortex and temporal parietal junction-and related coupling patterns to both objective negotiation behaviors and subjective experiences. While conventional measures of neural synchrony, such as inter-subject correlation and wavelet coherence, showed no relationships with outcomes, CRQA revealed systematic associations between dynamic coupling patterns and successful interaction. We found that balanced neural coordination, where leading and lagging flowed symmetrically between partners, predicted greater collaborative allocation adjustment and more positive social experiences, including shared understanding, cooperation, and liking. Longer delays in neurocognitive coordination, as opposed to immediate alignment, were associated with greater feelings of motivation during the interaction. Finally, greater complexity, or entropy, of neural coupling was linked to more parity in how partners moved toward the joint solution rather than a one-sided accommodation. These findings demonstrate that real social interaction can be captured through analytical methods that account for the dynamic, nonlinear processes being studied, creating new possibilities for understanding how minds connect during natural human interaction.