July 27, 2020

Collective Intelligence – does it naturally emerge or can it be cultivated intentionally?

By: Barbro Moen Ternsten

Generally speaking, collective intelligence is the intelligence of human groups as a whole, be those organizational teams, families, or even nations. There are many definitions for it. One of them is the following:

“Collective intelligence is when a group of agents perform a series and a combination of behaviors that together lead to the emergence of an intelligent collective response, i.e. more benefits than what would be obtained by single-handed individual efforts.”

The topic of collective intelligence, however, poses more questions than it answers. There is one question in particular that we need to elaborate on to reach a deeper understanding of collective intelligence: Does collective intelligence naturally emerge, or can it be cultivated intentionally? This question is particularly important because if collective intelligence is malleable, then we need to find it how. 

To answer this question, we need to pose even more questions: what evidence suggests that collective intelligence is something that emerges, and what evidence suggests that collective intelligence can be cultivated intentionally? Is emergence more likely than intention to cultivate the collective intelligence? Or is it the other way around? Will intention make emergence more unlikely? Getting answers to some of these questions will help us in our mission to improve group performance. 

What predicts collective intelligence?

In a research study performed by Woolley, Aggarwal and Malone (Collective Intelligence and Group Performance, 2015) the authors found strong support for the existence of a general collective intelligence factor (c). This factor predicts the performance of a group on a wide variety of tasks, and it was measured in groups while they performed a variety of demanding tasks. Further on, this factor was again used in further statistical regression analysis to assess whether it was predictive of the collective intelligence of the same groups but on completely different tasks. Remarkably, the analysis showed that at least twice as much variance in performance was predicted by collective intelligence than by individual intelligence (g). In other words, collective intelligence mattered for group performance more than individual intelligence did.

“As expected, we found that c (collective intelligence) was a significant predictor of group performance on both criterion tasks, and – surprisingly – the average individual intelligence of group members was not.”

Weak link between individual intelligence and collective intelligence

 A moderate correlation was found between the average and maximum intelligence of individual group members and collective intelligence. However, a much stronger predictor of collective intelligence was the average social perceptiveness of the group members, as measured by the Reading the Mind in the Eyes (RME) test (Baron-Cohen, Wheelwright, Hill, Raste & Plumb, 2001). This test measures people’s ability to judge others’ emotions from looking only at a picture of their eyes. Higher score on this test indicated that the group will likely have higher collective intelligence. 

Woolley, Aggarwal and Malone also found that the proportion of women in mattered for the collective intelligence of the group. The reason for this, however, was that women generally scored higher on tests like the RME than men. Thus, the collective intelligence of a group can be improved by adding people with high social perceptiveness (or social sensitivity). 

Diversity and collective intelligence

It is commonly believed that diversity is good for the collective performance of the group. However, generalization here is not wise, and the results depend on the task. Usually, groups performing creative or innovative tasks are likely to benefit from diversity, whereas groups performing tasks for which rapid delivery is important are often derailed by diversity (Williams & O’Reilly, 1998). 

Moreover, too much diversity can become too much of a good thing. Researchers have found a curvilinear, inverted U-shaped relationship between cognitive-style diversity and collective intelligence. In other words, while moderate diversity in cognitive styles is beneficial, extreme diversity can hurt the team (Aggarwal & Woolley, 2013). 

Bottom-up compositional features and top-down interaction 

These factors, namely social perceptiveness and cognitive diversity are compositional features of collective intelligence, and they may be somewhat difficult to control. The authors of the study also talked about the interaction processes in the team. Even though we are still far from a complete process theory that explains why some groups are more intelligent than others, it has been observed that more collective intelligent groups communicate more and participate more equally than other groups. For instance, several researchers have found that collective intelligence was significantly predicted by the total amounts of spoken communication in face-to-face groups and written communication in online groups (Engel et al., 2014; Kim et al., 2015; Woolley et al., 2010). These findings seem reasonable, since groups in which people communicate more and participate more actively and equally are more likely to be able to take advantage of the full knowledge and skills of all their members. 

Engel et al. also found that collective intelligence was predicted by how equally communication and work contribution were distributed among group members in both face-to-face and online groups. A group where one or two is dominating the activity are less collective intelligent than groups where activity is spread more equally. 


It is assumed that compositional features of a group (represented here by social perceptiveness and cognitive diversity), combined with the interaction processes (represented here by level and equality of communication and equality in work contribution) both will impact the group’s collective intelligence. We will argue that the compositional features are more difficult to cultivate intentionally than the interaction processes described here. Whereas a group with a set number of people will be difficult to change in terms of each group members individual social perceptiveness and cognition, it will be easier to adapt and change the processes like how team members interact with one another, what rules they decide for the group as a whole in terms of communication and work contribution, and how they intentionally learn and adjust along the way. 


Woolley, W. A., Aggarwal, I., & Malone , T. (2015). Collective Intelligence and Group Performance.

Barazi, M. (2020). Fostering Collectively Intelligent Teams. Retrieved from Vital Intelligence: https://www.vital-intelligence.com/2020/05/23/fostering-collectively-intelligent-teams/