The challenges facing the scientific community are numerous, and I’ve attempted to cover many of them on this blog over the years. With science getting ever more expensive however, attempts to boost efficiency are vitally important. Most of these attempts focus on the use of technology to support the scientific process, with technologies such as artificial intelligence used to find connections in seemingly disparate places, or crunch vast quantities of data quicker than ever before.
A new study from the University of Chicago highlights how we could also achieve a productivity boost by focusing on the makeup of scientific teams. The research analyzed over 65 million patents, research papers and software projects to try and identify the optimum team size for success.
The analysis found that small teams were nearly always better, and such teams were much more likely to introduce groundbreaking new ideas, with larger teams doing better work in developed fields.
“Big teams are almost always more conservative. The work they produce is like blockbuster sequels; very reactive and low-risk.” the researchers say. “Bigger teams are always searching the immediate past, always building on yesterday’s hits. Whereas the small teams, they do weird stuff—they’re reaching further into the past, and it takes longer for others to understand and appreciate the potential of what they are doing.”
Ideal teams
In total, the researchers mined through some 44 million research articles, 600 million citations, 5 million patents and over 16 million software projects posted on Github. The algorithm the team developed then assessed each piece of work to determine how disruptive it was to their respective field.
“Intuitively, a disruptive paper is like the moon during the lunar eclipse; it overshadows the sun—the idea it builds upon—and redirects all future attention to itself,” the authors explain. “The fact that most of the future works only cite the focal paper and not its references is evidence for the ‘novelty’ of the focal paper. Therefore, we can use this measure, originally proposed by Funk and Owen-Smith, as a proxy for the creation of new directions in the history of science and technology.”
Interestingly, whether it was academic research or software development, the disruptive capabilities of the project declined with every extra person added to the team.
The power of small teams
The researchers suggest that small teams were more disruptive in large part because of how they seem to view the history of their field. For large teams, they tended to look at highly cited research, and therefore ‘stand on the shoulder of giants’, whereas smaller teams went much further back and would often cite older and less popular ideas.
“Small teams and large teams are different in nature,” the authors explain. “Small teams remember forgotten ideas, ask questions and create new directions, whereas large teams chase hotspots and forget less popular ideas, answer questions and stabilize established paradigms.”
That’s not to say that large teams are bad of course, for they aren’t at all, just that they play a different role in the research ecosystem. Small teams are great for coming up with new and exciting insights, which are then advanced by larger teams.
The challenge is that research funders are tending to veer towards ever larger teams, which may hinder our efforts in creating big leaps forward in understanding. The researchers believe that a more effective approach would be to fund a diverse range of approaches via a VC-style approach.
“Most things are going to fail, or are not going to push the needle within a field. As a result it’s really about optimizing failure,” the authors conclude. “If you want to do discovery, you have to gamble.”
