This dissertation explores how group size affects collective intelligence. It is composed of
three quantitative studies. The first study explores how time pressure in small groups (size
4) and large groups (size 20) affected collective intelligence. The results showed that the
large groups significantly and consistently outperformed the small groups in different time
pressure conditions. This led to the second study which explored whether the collaboration
tool used in the first study might have provided unexpected benefits for large groups that
counteracted any process loss in the large groups. While the results from the second study
confirmed that the collaboration tool did indeed significantly improve the collective
intelligence score of groups, one surprising result was that this effect occurred, not only in
large groups (size 20), but also in small ones (size 4). The final study then set out to explore
this surprising result in further detail by including a variety of group sizes (sizes 5, 10, 15, 20,
25, 30, 35 and 40) in both the collaboration conditions. It was hypothesized that by including
more group sizes, the study would determine whether a curvilinear (inverted-U) relationship
existed. The results not only confirmed the curvilinear (inverted-U) relationship but also
suggested an optimal group size of about 30 for groups with the collaboration tool and 25
for groups without the collaboration tool.