Optimizing creative teams
An experiment in Rome singled out key determinants of efficient teams.
Modern communication technologies have made a completely new kind of creative work possible. Whether Wikipedia, coding new computer games, developing screenplays or songs—many creative processes today are the product of a crowd. Individuals only contribute chunks of content to the big picture.
Bernardo Monechi (Sony Computer Science Lab Paris), Giulia Pullano (INSERM at the Sorbonne in Paris) and CSH’s Vittorio Loreto wanted to know under which conditions such creative communities are most productive. The team designed an open-ended experiment to identify key ingredients of successful working groups.
Real life interactions
The scientists set up three stations with white Lego bricks at the KREYON Days exhibition in Rome. Each station had an assigned topic: “spring”, “pyramids” and “Halloween”. Visitors to the exhibition could freely contribute to the collective development of one, two, or all of the Lego artworks, either by building, destroying, or modifying some of their parts.
The scientists wanted to document time and pattern of the individual’s interactions as well as their contributions to a station. Therefor they equipped all participants of the experiment with special sensors when they entered the working area. These Radio Frequency Identification (RFID) sensors with a 20 seconds time resolution recorded an interaction of two when they stood in front of each other within one meter for at least 20 seconds. The sensors also documented the time spent at each of the tables.
“The constant monitoring shows us the dynamics of social bonds between people,” explains Vittorio. These social bonds continuously form and break away, leading to a constant restructuring of the working teams.
In parallel, infrared depth sensors monitored the evolution and the volume of each of the Lego artworks. This allowed accurate real-time 3D reconstructions of the growing and changing constructions.
In total, around 600 people participated in the experiment.
Keys to successful constructing: commitment and social influence
The researchers were able to identify some components of successful working teams. “Larger teams with committed people—that is, people focusing primarily on one station—are building more successfully,” says Vittorio. The higher the commitment and the bigger the team, the higher the growth rates.
“We also looked at the role of socially influential people in this dynamical interaction network,” Vittorio continues. Such influencers are defined as information spreaders within the social network of the experiment, analogue to virus spreaders during an epidemy. Influencers were identified through a simple information diffusion model, the SI (Susceptible-Infected) model. Every individual is associated with observable fractions of interaction. This allows the scientists to quantify his or her level of influence. When more influential people gathered at a time, the artworks grew faster.
New people bring in new ideas
On the other hand, it also seems to make sense to have a certain number of people who have not interacted much previously. The researchers call them “weak ties” in a given social network.
„Some experts argue, that ‚weak ties‘ in social networks are the most critical drivers of collective creativity,” says Vittorio. Weak ties contribute to a constant inflow of new information. Hence, too many weak ties could prevent an efficient communication between individuals.
“Our study shows that there exists an optimal fraction of weak ties in a working team that maximizes the building efficiency,” says Vittorio. This optimal ratio highlights a subtle interplay between the strategies of exploiting (retracing old steps) and exploring (experimenting novelties).
The scientists summon their observations up in one sentence: “The best way to assemble a team for a creative task is to have it large, full of firmly committed, possibly influential, individuals, and with a right balance between weak and strong ties,” the scientists write.
The study “Efficient team structures in an open-ended cooperative creativity experiment” appeared in October in the journal PNAS.
CSH Workshop: “Using Mobile Telephony Data for Societal Insight”
Dec 10, 2019 | 9:30—Dec 11, 2019 | 17:00
CSH Talk by Stefan Schmid: “Self-adjusting networks”
Dec 06, 2019 | 15:00—16:00
CSH Talk by Tiago Santos: “Self- and Cross-Excitation in Stack Exchange Question & Answers Communities”
Dec 04, 2019 | 15:00—16:00