Brainstorming: Interbrain coupling in groups forms the basis of group creativity

Creativity is often framed as an individual capability, yet it is fundamentally shaped by social context. Creative activities and evaluations typically occur in social settings and are influenced by cultural norms. Despite this, most research has examined individuals in isolation, with only emerging work on dyads; the neural mechanisms underpinning group creativity remain unclear. The present study treats the group as the primary unit of analysis and proposes two complementary group mindsets that influence creative outcomes: (1) herding/alignment (a tendency to blend in, synchronize, and imitate), which can facilitate cooperation but may reduce originality through imitation; and (2) flexibility/independence (openness to diverse ideas and exploration of alternatives), which underlies creative thought. The authors hypothesize two distinct neural networks supporting these mindsets and predict that they differentially contribute to group creativity (Hypothesis 1), and that higher group creativity is associated with a greater degree of flexibility relative to herding (Hypothesis 2). They motivate DLPFC as central to flexibility and executive control, and IFG as central to imitation and social alignment.

The study builds on a growing literature recognizing creativity as a socially embedded phenomenon that varies with cultural and group dynamics. Prior work highlights: (a) the role of herding/synchronization processes in social interaction, often implicating the IFG and mirror neuron system in imitation, coordination, and perception–action matching; (b) the involvement of executive control networks, especially the DLPFC, in cognitive flexibility, shifting, and decision-making supporting creative generation and selection; and (c) contributions from large-scale networks such as the Default Mode Network (DMN) for associative idea generation and the Executive Control Network (ECN) for controlled selection. Empirical studies in hyperscanning contexts have linked inter-brain synchrony to cooperation and coordination but raised questions about its impact on originality. The authors leverage this background to test whether IFG interbrain coupling (herding) relates negatively to creativity while DLPFC interbrain coupling (flexibility) relates positively, and whether the balance between these couplings best predicts group creative performance.

Design and participants: 22 groups (N=88; 12 males, 76 females; mean age 22.97 ± 4.28), each composed of four unfamiliar, right-handed participants, were recruited at the University of Haifa. Groups were gender-homogeneous. Ethics approval and informed consent were obtained. Participants received payment or course credit. Tasks and procedure: Groups performed two collaborative divergent creativity tasks while seated face-to-face: (1) Object Alternation Task (OAT/AUT-like divergent uses task), generating alternative uses for common objects in 30-second rounds; and (2) Egg task, generating ways to prevent an egg from breaking after a fall, within two minutes. Tasks were counterbalanced across groups. Each round began with individual suggestions to ensure participation, followed by free group brainstorming. Verbal responses were audio-recorded and transcribed. Creativity assessment: Expert raters scored flexibility (number of categories) and originality (rarity-weighted scoring). Frequency thresholds: >10% frequency = 0 points; 5–10% = 1 point; <5% = 2 points. Originality and flexibility were computed per task; total creativity per task was the mean of originality and flexibility; final group creativity score was the sum across the two tasks. Individual creativity was assessed with a TTCT figural subset (circle-based drawings), scoring originality (rarity-weighted per TTCT guidelines) and flexibility (category count across 70 categories, including two added categories). Total individual creativity was the mean of originality and flexibility. Herding assessment: A trained rater quantified repeated or conceptually similar responses within groups during AUT items. Verbatim repetitions received 1 point; conceptually similar but differently phrased responses received 0.5 points. The herding score was the mean across responses. fNIRS acquisition: Hyperscanning with Brite24/Artinis fNIRS measured O2Hb and HHb at 760/850 nm, sampling at 50 Hz, using 18 sources and 8 detectors (24 channels), positioned per EEG 10–20 system. ROIs included left/right IFG (pars opercularis/triangularis approximations) and left/right DLPFC (BA46/BA9). Preprocessing: Implemented in HOMER3 with a five-step pipeline: (1) convert light intensity to optical density; (2) detect and correct motion artifacts via targeted PCA; (3) bandpass filter (0–1 Hz); (4) convert to O2Hb/HHb via modified Beer–Lambert law; (5) remove channels with positive O2Hb–HHb correlations >0.5 (expecting negative coupling), to improve signal quality. Analyses focused primarily on O2Hb as more sensitive to cerebral blood flow changes. Interbrain coupling computation: For each ROI and for each dyad among the four participants (AB, AC, AD, BC, BD, CD), Wavelet Transform Coherence (WTC; Morlet wavelet, 0.015–0.15 Hz) was computed between O2Hb time series using MATLAB (wavedec coherence package). WTC values range 0–1. Group-level interbrain coupling per ROI was the average WTC across all six dyads. Pseudo-group analyses were constructed by pairing individuals from different groups to estimate chance-level coupling. Interbrain coupling estimates were averaged across the two tasks to form group-level coupling indices. Statistical analysis: A mixed-design repeated-measures ANOVA tested effects of group (real vs pseudo; between) and brain area (IFG vs DLPFC) and hemisphere (left vs right; within) on interbrain coupling. One-way MANOVA/paired comparisons examined region-specific real vs pseudo differences. Repeated-measures ANOVAs within real and pseudo groups compared regions. Pearson correlations probed associations between interbrain coupling and creativity. Stepwise linear regressions predicted (a) herding and (b) group creativity from regional interbrain coupling. Additional regressions tested coupling ratios (e.g., L.DLPFC/R.IFG) as predictors of herding and creativity. Linear mixed-effects models tested whether interbrain coupling predicted individual creativity (random intercepts for group).

- Real vs pseudo groups: Interbrain coupling was higher in real than pseudo groups (ANOVA main effect of group: F(1,42) ≈ 23.22, p < 0.001, partial η² = 0.36; real M = 0.32, SD = 0.01 vs pseudo M = 0.30, SD = 0.01). - Brain area effects: DLPFC showed higher interbrain coupling than IFG (F(1,42) ≈ 16.55, p < 0.001, partial η² ≈ 0.28). A significant three-way interaction of group × area × hemisphere was observed (F(1,42) ≈ 4.27, p < 0.05, partial η² = 0.09). - Region-wise real vs pseudo differences: Interbrain coupling was higher in real groups across all regions: R.IFG, L.IFG, R.DLPFC, L.DLPFC (all p < 0.001). - Within real groups, regional differences: Repeated-measures ANOVA significant for real groups (F(3,319) = 5.77, p < 0.001, partial η² = 0.48) but not for pseudo groups. Post hoc tests showed L.DLPFC > L.IFG (p < 0.05) and L.DLPFC > R.IFG (p < 0.01). - Herding prediction: Higher interbrain coupling in R.IFG predicted higher herding (F(1,20) = 8.61, p < 0.01; adjusted R² = 0.272). - Creativity prediction (regional couplings): Stepwise regression indicated L.DLPFC positively predicted group creativity (β = 0.384, p < 0.001), whereas R.IFG (β = −0.42, t = −2.64, p < 0.005) and L.IFG (β = −0.43, t = −2.224, p < 0.005) negatively predicted creativity. - Coupling ratios: The L.DLPFC/R.IFG ratio predicted lower herding (F(1,20) = 6.73, p < 0.005; adjusted R² = 0.21) and higher creativity (β = 0.63, t = 3.25, p < 0.01; F(1,20) = 13.32, p < 0.01; adjusted R² = 0.37). - Individual creativity: LME analyses showed no significant effects of interbrain coupling in L.DLPFC, R.IFG, R.DLPFC, or L.IFG on individual creativity (all p > 0.27).

Findings support the hypothesis that distinct group mindsets—herding and flexibility—map onto IFG and DLPFC interbrain coupling, respectively, and differentially relate to group creativity. Elevated IFG interbrain coupling, particularly in right IFG, was associated with greater herding/imitative responding and lower originality, consistent with the IFG’s role in synchronization, imitation, and perception–action matching. This suggests that strong social alignment may constrain divergent thinking during brainstorming. In contrast, increased interbrain coupling in DLPFC, especially left DLPFC in this verbal divergent context, was positively related to group creativity, consistent with DLPFC’s role in executive control, cognitive flexibility, and the selection/shifting among ideas. Crucially, creativity was best explained by the balance between flexibility and herding networks, quantified by the L.DLPFC/R.IFG coupling ratio: higher balance (greater relative DLPFC coupling) predicted higher creativity and lower herding. The absence of relationships between coupling and individual creativity underscores that group-level neural coordination, rather than individual ability, primarily drives collective creative outcomes in interactive contexts. These results align with models positing that optimal creativity requires coordinated engagement of associative (DMN) and control (ECN) processes, with social interaction dynamics modulating the balance via IFG and DLPFC coupling.

This study provides evidence that group creativity emerges from the interplay between two group-level neural dynamics: herding (IFG coupling) and flexibility (DLPFC coupling). During brainstorming, real interacting groups exhibited greater interbrain coupling than pseudo groups across IFG and DLPFC, with left DLPFC coupling especially elevated. Importantly, DLPFC coupling positively, and IFG coupling negatively, predicted group creativity, and their ratio (L.DLPFC/R.IFG) best explained variance in both creativity (positive) and herding (negative). These findings establish interbrain coupling as a neural substrate of group mindsets shaping collective creative performance and advocate fostering a flexible mindset while limiting excessive synchronization to enhance group creativity. Future work should test these mechanisms across more diverse samples, manipulate interaction structures to modulate flexibility vs herding, explore different task modalities and durations, and integrate broader network measures (e.g., DMN–ECN interactions) to refine neurocognitive models of group creativity.

- Sample composition was predominantly female, which may limit generalizability and is pertinent given reported gender differences in creativity and neural correlates of creative thinking. - Task duration was relatively short (~8 minutes), leaving open questions about temporal dynamics of interbrain coupling and creativity over longer interactions. - The study focused on specific ROIs (bilateral IFG and DLPFC) and O2Hb signals; broader network coverage and multimodal measures may reveal additional mechanisms. - Creativity tasks emphasized verbal divergent thinking; findings may not fully generalize to nonverbal or convergent creative tasks.