Evolutionary emergence of collective intelligence in large groups of students

Frontiers in Psychology

Psychology

Santos Orejudo¹  Pablo Bautista¹  Ana Belén Cebollero-Salinas²  Jacobo Cano-Escoriaza²  Pilar Rivero³  Jesús Clemente-Gallardo⁴  Alfonso Tarancón⁴  Alejandro Rivero⁵ 

[1] Department of Psychology and Sociology, University of Zaragoza, Zaragoza, Spain;Department of Sciences of Education, University of Zaragoza, Zaragoza, Spain;Department of Specific Didactics, University of Zaragoza, Zaragoza, Spain;Department of Theoretical Physics, University of Zaragoza, Zaragoza, Spain;Institute for Biocomputation and Physics of Complex Systems, University of Zaragoza, Zaragoza, Spain;Kampal Data Solutions, Zaragoza, Spain;

关键词: collective intelligence;    social experiment;    web platform;    mathematics test;    moral reasoning;   

DOI  :  10.3389/fpsyg.2022.848048

 received in 2022-01-04, accepted in 2022-09-26,  发布年份 2022

来源: Frontiers

【 摘 要 】

The emergence of collective intelligence has been studied in much greater detail in small groups than in larger ones. Nevertheless, in groups of several hundreds or thousands of members, it is well-known that the social environment exerts a considerable influence on individual behavior. A few recent papers have dealt with some aspects of large group situations, but have not provided an in-depth analysis of the role of interactions among the members of a group in the creation of ideas, as well as the group’s overall performance. In this study, we report an experiment where a large set of individuals, i.e., 789 high-school students, cooperated online in real time to solve two different examinations on a specifically designed platform (Thinkhub). Our goal of this paper 6 to describe the specific mechanisms of idea creation we were able to observe and to measure the group’s performance as a whole. When we deal with communication networks featuring a large number of interacting entities, it seems natural to model the set as a complex system by resorting to the tools of statistical mechanics. Our experiment shows how an interaction in small groups that increase in size over several phases, leading to a final phase where the students are confronted with the most popular answers of the previous phases, is capable of producing high-quality answers to all examination questions, whereby the last phase plays a crucial role. Our experiment likewise shows that a group’s performance in such a task progresses in a linear manner in parallel with the size of the group. Finally, we show that the controlled interaction and dynamics foreseen in the system can reduce the spread of “fake news” within the group.

【 授权许可】

Unknown   
Copyright © 2022 Orejudo, Cano-Escoriaza, Cebollero-Salinas, Bautista, Clemente-Gallardo, Rivero, Rivero and Tarancón.