What makes a successful research team?

The availability of online databases containing millions of papers, as well as analytical tools from network science have turbo-charged studies of collaboration patterns among researchers. These studies typically try to answer the question: "How do you assemble a network and figure out where to plug in your ideas to get the best return?"

A recent article by John Whitfield in Nature summarizes the researchers' tentative tips on what their work reveals:

• If you team up with someone from another institution (of equal or higher tier to your own), the resulting papers are more highly cited than if you team with someone from your own institution.

• If you specialize and work mostly with people in your own field, or you become a big generalist and work with people in a wide range of fields, you get highly cited papers. But in the middle (people who work with an intermediate number of other fields) you get less successful papers.

• In large interdisciplinary groups, the more diverse they were, the less productive they were, but groups where the authors had previous papers together were much more successful than others.

• Large collaborations (say a team of around 20) were more successful if they had a high rate of turnover, but small groups (say 3 to 4) were more successful if they were stable. But stable small groups tend to publish lower impact papers over time.