People tend to naturally coordinate their body movements more smoothly when interacting with people of similar age or ethnic background than when speaking with people of different backgrounds. These physical rhythms often occur without people being aware of them, suggesting that common, visible characteristics provide the basis for smoother interpersonal relationships. This research recently Personality and Social Psychology Bulletin.
Spontaneous behavioral synchronization refers to the way people unconsciously adjust their actions, movements, and pace during a conversation. This phenomenon involves moving at the same time, matching the speed of movements, and smoothly blending gestures. It looks like a subtle, unchoreographed dance between two conversation partners.
In psychological research, this physical mirroring typically indicates relationship compatibility and rapport, a sense of intimacy between individuals. When two people click, their bodies often move in unison, reflecting their connection. Scientists wanted to understand the specific factors that predict whether this physical adjustment will emerge naturally between strangers.
“Much of the existing research asks what happens after people become behaviorally aligned; for example, does it lead to greater trust, stronger relationships, or better teamwork?” said Morgan D. Stosic, a research psychologist at NASA’s Johnson Space Center KBR. Stosic, who earned his doctorate in psychology from the University of Maine in 2023, conducted this research separately from his job at NASA. “There has been less attention to the more fundamental question of what makes people conform in the first place.”
Past theories suggest that people who perceive themselves to be similar to each other may be more relaxed and experience greater chemistry. Specifically, highly visible characteristics such as age and race are some of the first characteristics people use to assess similarity. Sharing these traits can reduce initial anxiety, free up mental energy, make it easier to focus on your partner, and make it easier to fall into a common physical rhythm.
“Our goal was to understand whether shared visible characteristics with others, particularly age and race, influence the natural connections people make,” Stosich said. “We wanted to better understand whether these common characteristics shape our social interactions from the first meeting.”
To test this idea, the researchers set up interactions between people who did and did not share the same race or age category. The research team evaluated three separate studies involving female participants who interacted in pairs known as dyads. Because women generally have higher rates of spontaneous behavioral adjustment than men, the authors focused strictly on women to hold gender constant.
In the first study, 182 college students created 91 pairs of strangers. These participants completed a problem-solving task collaboratively via a video conferencing platform. Some pairs shared the same ethnic background, while others did not.
“We looked at two results: behavioral synchrony, the subtle tendency for people’s movements and actions to adjust over time, and how closely connected people reported feeling after meeting each other,” Stosich told SciPost. “To study this, we paired strangers of similar or different ages and races and observed how they interacted.”
The scientists recorded the interaction and extracted three separate one-minute video clips from the beginning, middle, and end of the conversation. Trained human programmers viewed these silent clips and rated the pairs for simultaneous movement, tempo similarity, and overall smoothness. The researchers also analyzed the same video clips using an automated software program that tracks the joints of the human body and measures simultaneous movements.
After the video task, participants filled out a questionnaire assessing how much trust they felt with their partner. The authors found that pairs who shared the same race synchronized their behavior to a greater extent than pairs with different racial backgrounds. This difference in movement coordination was only detected by human programmers.
The automated software did not record any differences based on race. The computer scores also showed no correlation with participants’ own ratings of trust, indicating that the software may not capture the essence of interpersonal relationships. Interestingly, even though same-race pairs moved more synchronously, participants themselves reported that they perceived no difference in trust compared to mixed-race pairs.
“One of the most interesting findings was that people did not report feeling closer to someone simply because they were the same age or race,” Stosich said. “However, trained observers rated people of similar age and race to move and interact in more synchronized ways. This suggests that subtle aspects of our behavior may reflect common traits, even if consciously or explicitly reported emotions do not.”
In the second study, we moved the interaction to a face-to-face laboratory setting to see if our findings held up in a face-to-face setting. The study involved 88 women in 44 pairs from a group of young adults aged 18 to 24 and older adults aged 60 to 90. The pair sat side by side and spent 15 minutes planning a pretend vacation together using a world map and fake money.
Again, the researchers categorized the pairs based on race and whether they matched in age. They used human coders and automated software to assess short video clips of interactions and collect self-reported rapport surveys. In this small sample, the scientists found no statistically significant differences in physical adjustment or self-reported trust based on race or age similarity.
To improve understanding in a larger group, the authors conducted a third study that replicated the in-person vacation planning task from the second study. This last group included 182 women making up 91 pairs. They performed the exact same task, but for 10 minutes instead of 15 minutes.
This third study found that same-age pairs synchronized their behavior to a greater extent than mixed-age pairs. As in the first study, this physical adjustment was recognized by human programmers rather than computer software. In this particular group, the effect of racial similarity on physical adjustment was not statistically significant.
To understand the different results across the three experiments, the researchers pooled all the data and conducted a combined analysis on a total of 450 people. By combining the data, we gained sufficient statistical power to identify broader trends. Combined analyzes showed that both same-age and same-race pairs exhibited significantly higher levels of behavioral synchrony than pairs of different backgrounds.
This combined analysis also confirmed that human observers were much better at detecting this adjustment than computer software. Human ratings were consistently consistent with the psychological concept of trust, providing evidence that the programmers were measuring meaningful social behavior. The automated software was unable to establish any trust relationships across the dataset.
“One of the biggest surprises was that the computer vision methods we tested didn’t work as well as we expected,” Stosic said. “Traditionally, behavioral synchrony has been measured by trained human observers, but researchers are increasingly turning to computer vision, which can analyze interactions more quickly.”
The automated tools used in this study only tracked joint joint movements, missing more subtle aspects such as shared tempo and overall smoothness that human observers can perceive. The video only showed the participants from the waist up, so the software may have had a hard time mapping their movements accurately.
“However, in our study, the automated approach was not able to accurately capture synchronization in the way we expected,” Stosic said. “That could be due to factors such as video quality, lighting, or limitations in the software itself. Whatever the reason, it was an important reminder that new technologies are not automatically better and that researchers need to carefully validate them before relying on them to measure complex human behavior.”
Another complicating finding is the gap between participants’ behavior and what they report in the survey. The combined data confirmed that age and racial similarity had no effect on the degree of compatibility felt by participants in the study.
“One important caveat is that people’s self-reported emotions don’t necessarily match their observable behaviors,” Stosich says. “While participants reported feeling similar closeness to their partners regardless of age or race, trained observers observed greater behavioral synchrony between people who were more similar.”
Stosich pointed out that this discrepancy doesn’t necessarily mean that one or the other measure is wrong. “Rather, it suggests that human connections are complex,” she says. “How close we feel to someone and how we act around that person may capture different aspects of social interaction.”
People in mixed-race or mixed-age pairs may have deliberately tailored their survey responses to avoid bias. Alternatively, a trusting relationship consists of multiple components, such as mutual attention and positivity. People in the mixed group may have compensated for the lack of physical rhythm by smiling more or paying more attention to their partners.
“While it’s possible that self-reports don’t necessarily fully reflect people’s true feelings, it’s equally possible that behavioral conformity is just part of what makes us feel connected to others,” Stosich says.
One limitation to consider is that the human programmers may have had unconscious biases and thought that pairs that were similar to each other were more harmonious. To test this, future studies could use video filters that obscure the participants’ physical identities, leaving only their outlines visible to the coder. Furthermore, the use of only female participants means that these findings may not apply to men or mixed gender groups.
“That’s where we want to see this research go,” Stosich said. “While our study focused on age and race, there are many other characteristics, both physical and psychological, that can influence whether people naturally ‘click’. A deeper understanding of what fosters these subtle social connections can provide greater insight into how successful relationships and effective teams develop in everyday life, the workplace, and even in high-stakes environments such as medicine and space flight. ”
The study, “Multi-method assessment of spontaneous behavioral synchrony in race- and age-matched versus mismatched dyads,” was authored by Morgan D. Stosic, Adele E. Weaver, Ken Fujiwara, Ishabel M. Vicaria, Derek M. Isakowitz, and Molly A. Reuben.
