Virtual social interactions during the COVID-19 pandemic: the effect of interpersonal motor synchrony on social interactions in the virtual space

The COVID-19 pandemic dramatically altered social interaction patterns, necessitating increased reliance on virtual communication. While the link between motor synchrony (spontaneous or intentional synchronization of movements) and emotional alignment (emotional contagion) has been extensively researched in face-to-face settings, its presence and effects within virtual interactions remained unclear. Prior research consistently demonstrates the pro-social effects of motor synchrony and emotional alignment in in-person interactions, including increased liking, trust, cohesion, and self-other overlap. However, studies involving virtual avatars did not consistently replicate these findings. The current study addresses this gap by exploring whether these phenomena occur during virtual interactions mediated by audio and video, and whether similar pro-social effects are observed. Understanding these aspects is crucial for developing effective strategies to counter the negative consequences of social distancing and enhance social connection in virtual spaces.

Extensive literature supports a strong link between motor synchrony and emotional alignment during face-to-face interactions. This connection is often attributed to shared neural mechanisms involving an error-monitoring system, an alignment system, and a reward system, creating a feedback loop where activation of one triggers the other. Studies consistently show that motor synchrony enhances prosocial behavior, liking, trust, and cohesion. However, inconsistent results exist in virtual settings, particularly when interactions involve virtual agents, questioning the generalizability of these findings to virtual human-human interactions. The mirror neuron system and action-observation network play a key role, enabling individuals to anticipate others' intentions and share bodily states. Embodied simulations, arising from these neural mechanisms, facilitate understanding and alignment of behaviors and emotions, contributing to empathy and feelings of closeness. The rise of virtual interactions due to the pandemic highlighted the need to examine if core social phenomena, such as spontaneous movement synchronization and emotional alignment, are maintained in such contexts, and whether they promote social connectedness.

This study involved 196 Hebrew-speaking university students (aged 19–30) randomly assigned to 98 same-sex, stranger dyads. Participants engaged in a 5-minute virtual interaction via Zoom, instructed to share difficulties experienced during the pandemic. To assess motor synchrony, Motion Energy Analysis (MEA) software objectively quantified movement synchronization from video recordings. MEA converts video pixels to grayscale, tracking changes in “energy” (differences between frames) within manually defined regions of interest (ROIs) representing each participant's head and shoulders. A time-lagged cross-correlation algorithm analyzed the synchrony of these time series. A pseudo-synchrony analysis, involving shuffled time series, served as a control for chance synchrony. Emotional alignment was measured using the Positive and Negative Affect Schedule (PANAS), calculating difference scores for positive and negative affect before and after the interaction (lower difference indicating higher alignment). Pro-social effects (trust, liking, cohesion, self-other overlap, perceived similarity) were assessed using established self-report measures before and after the interaction. Data analysis involved t-tests to validate the emergence of synchrony and emotional alignment, MANOVA to compare pro-social scores before and after the interaction, and regression analysis to explore the relationship between synchrony, emotional alignment, and changes in pro-social effects.

The study successfully validated its task, demonstrating that motor synchrony and emotional alignment arose spontaneously during virtual interactions. Real dyads exhibited significantly higher synchrony (M = 0.14, SD = 0.04) than pseudo-dyads (M = 0.04, SD = 0.02), t(9727) = 39.92, p = 0.000. A significant decrease in the difference in negative affect scores (higher alignment) was observed after the interaction (Mbefore = 6.50, SD = 4.74; Mafter = 4.01, SD = 4.39), t(68) = 4.12, p = 0.000. MANOVA revealed significant increases in positive affect, trust, liking, cohesion, self-other overlap, and perceived similarity after the interaction (all ps < 0.001). Regression analysis showed that higher synchrony significantly predicted increased positive emotional alignment and liking (β = 0.394, p < 0.001; β = 0.496, p < 0.001 respectively). These findings suggest that virtual interactions, similar to face-to-face interactions, can induce pro-social effects, with synchrony playing a key role in positive emotional alignment and liking.

The findings support the hypothesis that motor synchrony, emotional alignment, and pro-social effects can occur during virtual social interactions. The reduction in negative affect despite sharing difficult experiences highlights the positive impact of shared emotional experiences, even virtually. This aligns with research showing the benefits of acknowledging shared emotional experiences. The study's results suggest that virtual interactions (using audio and video) can provide a similar opportunity for connectedness as face-to-face interactions. This supports the idea that motor synchrony and emotional alignment are underlying mechanisms of this connection-promoting effect. The study provides evidence against claims that virtual interactions are inherently less capable of fostering social connection. However, it also notes a discrepancy with studies that failed to find similar effects in virtual human-agent interactions, highlighting the importance of the interaction's specific nature and the types of virtual interaction being studied (audio and video versus text).

This study demonstrates for the first time the spontaneous emergence of motor synchrony and emotional alignment in virtual human-human interactions, leading to increased pro-sociality. The specific paradigm used yielded outcomes similar to face-to-face interactions, but also unique aspects emphasizing the multifaceted nature of interaction dynamics. The findings provide a foundation for developing new intervention protocols addressing social isolation, suggesting that carefully designed virtual interactions can be therapeutic and promote social connection. Future research should examine diverse virtual interaction types and explore the causal relationships between synchrony, emotional alignment, and pro-social outcomes.

The study lacked a comparison group, precluding direct comparison with face-to-face interactions. The design doesn't establish causality between synchrony, emotional alignment, and pro-social effects. Sample demographics (age range, gender ratio, university affiliation) might limit generalizability. The narrow ROI in MEA might have missed other body movement synchrony. The presence of a research assistant, although minimized, could have influenced interactions. Future studies should address these limitations.