Collective Intelligence Platforms – Does Communication Improve Performance Under Uncertainty?

Abstract

This paper reports the results of an experiment that investigates how collective communication on an online platform influences the quality of predictions of uncertain future events and thus the collective intelligence of a crowd in situations of uncertainty. Specifically, individual performance in predicting future events without communication is compared with performance after collective communication. Moreover, the paper investigates the effects of communication intensity on the change of prediction performance and explores the role of knowledge and intrinsic motivation. The findings show that topic-specific knowledge has a positive impact on prediction accuracy. Furthermore, the results support the expected positive effect of collective communication on online platforms on prediction accuracy. Additionally, the findings confirm an indirect effect of communication intensity on the improvement of prediction accuracy via knowledge gain. The influence of intrinsic motivation, however, depends on the topic of prediction. The results complement existing research on design factors of collective intelligence platforms by underscoring the effects of collective communication in large groups and by offering insight into the underlying processes. The results also have important practical implications for the design of collective intelligence platforms in situations of uncertainty. They highlight the need for collective communication to foster high prediction performance. Furthermore, the role of topic-specific knowledge is illustrated and should be considered when designing online platforms.