While the "wisdom of the crowd" shapes the behavior of large groups of people, less is known about the dynamics of small groups and how people interact to make decisions, particularly with respect to the behavior of individuals. Emergence of leaders, a key area of investigation in research organizations. The phenomenon is essential to understand social networks of all kinds.
Today, researchers at the NYU Tandon School of Engineering have deciphered the code on how leaders train with small groups of people over time. The work is detailed in a study entitled "Social Information and Spontaneous Emergence of Leaders in Human Groups", published in the Royal Society. Interface.
The team consisted of Maurizio Porfiri, Professor in the Department of Mechanical and Aerospace Engineering and Biomedical Engineering at NYU Tandon; Shinnosuke Nakayama, postdoctoral researcher at NYU Tandon; Elizabeth Krasner, an undergraduate student at NYU Tandon; and Lorenzo Zino, a Ph.D. visiting student at NYU Tandon of Politecnico di Torino, Italy.
To conduct the research, the team gathered several groups of five volunteers to participate in a cognitive test organized in 10 consecutive rounds. The task was to estimate the number of points displayed for half a second on a large screen. At each turn, participants were asked to choose one of the answers with the help of a personalized selector, without verbally communicating with each other. Because the dots were visible only for a moment, the band members, not having the time to count them, had to guess. However, the experiments were structured so that participants could modify their responses based on the responses of other members of their group: once all participants had chosen their initial responses, the screen – visible to all – displayed the current responses of all members with their past. performance in the selection of correct answers. Participants then had a 10-second window to modify their responses based on those of other group members.
The researchers, analyzing the evolution of participants' responses during the experiment, found that individuals did not choose the simple majority rule, as advocated by the wisdom of the crowd. Instead, they dynamically decided who to follow in their decisions, depending on the performance of each member of the group over time. On the basis of their observations, the researchers deduced a dynamic evolution of the interaction network, in which the participants were nodes and the links were the consequences of the social influence. For example, the investigators generated a link from one participant to another if the former had changed his answer to that of the second. The speed at which the network has grown has increased during each of the rounds.
"People used social information more and more over time, and the more accurate the information was, the more they affected the choices of participants," said Porfiri. "As a result, the relationship between participants' performance and their social influence has increased over time, which has resulted in the emergence of group leaders."
To discern the influence of social networks within the evolving group dynamics, investigators noted that:
- Participants were influenced by social information by changing their responses. On average, participants changed their response to an answer that no one had selected about 5% of the time. (There were more than 9 occurrences in all rounds in which participants changed their responses to those of others and only 0.6 in which they changed in response to those that no one else has. had selected.)
- Participants were more likely to be copied by others if their performances were good, even though their responses differed from those of the majority of the group.
Nakayama, the main author, explained that the behavior of small groups is extremely different from that of much larger gatherings.
"Where a large crowd would adopt a simple majority rule, with increased precision of performance over repeated interactions, individuals rely more on social information than on personal information, and as a result, good performers would emerge as other group leaders over time, "he said.
Porfiri noted that while most studies of social connections are based on static networks, defined by established established relationships, his team's research has focused on functional networks, based on ever-changing connections.
"Historically, social networks were equated with anatomical networks based on static links, and we suggest that, in small groups, networks evolve over time depending on actions." Actions based are like networks. in the brain, where physically distant neurons make connections to a specific function. "Our approach is analogous to learning neural circuits based on their functioning in the brain, rather than on how they are connected anatomically."
Sharing of information interferes with the "wisdom of the crowds": study
Shinnosuke Nakayama et al, Social Information and Spontaneous Emergence of Leaders in Human Groups, Journal of the Royal Society Interface (2019). DOI: 10.1098 / rsif.2018.0938