Julija Mell and her collaborators asked 112 teams, each comprising three students, to rank the commercial prospects of five different drink products from best to worst. Each member read some unique information themed by specialism: for example, one it was research and development (R&D) data about the five products; for another, information about legal and marketing aspects. The task had an ideal answer, and good performance required seeing interdependencies – for example, a chemical used to manufacture one product (R&D data) was at risk of being outlawed (legal).
In half of these teams, one member was given a written overview of the specialties held by the various members. In the remaining teams, this information was divided across members, so A might know B possesses legal info (but not that they also possess marketing info), B knows what C possesses, and so on. The take-home result? After their 15 minutes of team discussion, those with one member "in the know" about member specialities produced better rankings compared to teams with divided metaknowledge. Why?
Previously, this research group from the Rotterdam School of Management has found that people are most ready to share information and spark debate when their attention is drawn to how we each possess very different knowledge and experiences, meaning that any single perspective is bound to be partial and incomplete.
They reasoned that in the current study, giving one individual all of a group's metaknowledge gave this person a good shove into this state, so that they took steps to uncover the submerged information; even better, this person’s modelling of such behaviour likely sparked the same in return from their team-mates. Evaluating videos of the team discussions confirmed this analysis - members of teams with one central meta-knowledge individual asked more questions of each other about who knew what, and went on to incorporate and fuse information together more often, and this contributed to their better performance.
Note, the strategy of centralising meta-knowledge is not always advantageous. A single centralised focus didn't help teams in another condition, where each member received all relevant information (legal, R&D) about one product, such as the "Health-Conscious Adults" drink. By their nature, these members are already interconnecting types of knowledge, and performed highly with or without a catalyst figure.
Sharing of meta-knowledge matters most when knowledge is segmented by speciality - all too familiar to those of us who have worked in silo'd organisations, and inevitable for some types of high-level decision-making. In real life, the chasm between the knowledge possessed by different specialists is likely to be even deeper than in the simulated conditions of this study.
In an ideal world, of course everyone would have a complete and shared understanding of who knows what. But in practice, it seems advantageous (and more realistic) to aim to have at least one individual - not necessarily a leader - with a strong overview of this information. Such individuals naturally catalyse knowledge-seeking and sharing across the team, which is bound to be more productive than top-down attempts to build team-wide meta-awareness through data living/hiding on intranets or through time-consuming update meetings._________________________________
Mell, J., van Knippenberg, D., & van Ginkel, W. (2014). The Catalyst Effect: The Impact of Transactive Memory System Structure on Team Performance Academy of Management Journal, 57 (4), 1154-1173 DOI: 10.5465/amj.2012.0589
Post written by Alex Fradera (@alexfradera) for the BPS Research Digest.