Collective intelligence is the measurable general capacity of a group to perform well across a diverse range of tasks, distinct from the average or peak intelligence of its individual members. Known as the c factor, it is driven primarily by members' social sensitivity, balanced conversational turn-taking, and the proportion of women in the group.
The c factor was identified by Woolley and colleagues through a series of experiments in which groups completed tasks spanning brainstorming, moral reasoning, visual puzzles, and negotiation. It explains approximately 43% of the variance in group task performance and predicts success on novel complex tasks at r=0.52 1, far outperforming both average individual IQ (r=0.15) and maximum individual IQ (r=0.19) as predictors of what a group will achieve.
The single strongest individual-level predictor of collective intelligence is average social sensitivity among group members, measured via the Reading the Mind in the Eyes test. This association holds in face-to-face and online groups alike, indicating it is not an artefact of in-person social cues 12. Groups in which one or two members dominate speaking time score markedly lower on the c factor, regardless of those speakers' individual intelligence 1.
The Transactive Systems Model of Collective Intelligence describes how group-level performance emerges from three collective cognitive processes: transactive memory (who knows what within the group), transactive attention (co-ordinated focus on relevant problems), and transactive reasoning (building systematically on each other's contributions) 4. A meta-analysis across 22 studies and 1,356 groups confirmed that collaboration process quality, not member composition alone, drives collective intelligence, a finding that generalised across task types, cultures, and group sizes 3.
Collective intelligence — well-connected, diverse groups can solve problems no individual member could.
A cross-functional team assembles to address a supply-chain disruption. One senior member dominates the early discussion; quieter specialists hold back. The team's solution converges on a narrow set of options already considered by the dominant speaker. A second team, facilitated to distribute speaking time equally and draw out contributions from every role, surfaces three viable alternatives the first team never identified.
The process structure, not the expertise present, determined which team solved the problem.
Collective intelligence determines whether a team will perform consistently across varied challenges or merely match the output of its strongest member in narrow domains. Because the c factor predicts performance at r=0.52 on novel tasks 1 while average member IQ predicts it at r=0.15, selecting teams by individual credentials alone misses most of the variance in what those teams will actually deliver 3.
For leaders, the practical significance is that collective intelligence is trainable. Structuring meetings to distribute speaking time, building conditions of psychological safety that allow reticent members to contribute, and selecting for social sensitivity when forming teams all target the most modifiable c-factor predictors 14. Remote teams show no measurable decline in collective intelligence relative to co-located groups when social sensitivity among members remains high, confirming that virtual collaboration does not inherently diminish a group's intellectual capacity 2.
Collective intelligence is a group-level trait distinct from the average intelligence of its members. It predicts group task performance at r=0.52, compared to r=0.15 for average member IQ. A team of moderately able individuals with high social sensitivity will typically outperform a team of high-IQ individuals who fail to distribute conversational participation.
The two most reliably modifiable predictors are equal distribution of speaking time across group members and high average social sensitivity. Leaders can raise collective intelligence by structuring meetings to prevent conversational dominance and by creating the psychological safety that enables quieter members to contribute substantively.
Remote and co-located teams show no significant difference in collective intelligence when social sensitivity among members remains high. The association between social sensitivity and the c factor holds equally across face-to-face and online collaboration settings, indicating that virtual work does not inherently reduce a group's intellectual capacity.
Evidence supports several leader-controllable levers. Distributing speaking time, building psychological safety, and selecting for social sensitivity when forming teams all target the most modifiable c-factor predictors. The Transactive Systems Model further suggests that cultivating transactive memory (clarity about who knows what) improves the reasoning quality that underpins group performance.
Why Incompetence Feels Like Competence: The Dunning-Kruger Effect Examined
Applied Flow Protocols: Domain-Specific Systems for Reliable Peak Performance
Burnout Test: Where Are You on the Burnout Spectrum Right Now?
90-Day Sleep Optimisation Protocol: Rebuild Your Recovery From the Ground Up
Digital Detox Science: What Actually Happens When You Block Algorithmic Feeds
The Psychology of Power: What Happens to the Brain When You Gain Authority
Cognitive Fuel: The Evidence-Based Nutritional Framework for Brain Performance
Network Intelligence: The Science of Strategic Relationship Building for Career Growth
The 90-Day Kickstarter Protocol
Your day-by-day reset for sleep, stress & energy · PDF