Keith Sawyer reports some more research on the limits of brainstorming. In short you get better results if you ask participants to work alone before working together.
And then there’s this:
In the brainstorming group, ideas were more likely to build off of other people’s ideas. In fact, that’s one of the standard guidelines that are given to a brainstorming group, to listen and build on other’s ideas. However, the researchers found that ideas that built on other ideas were, on average, of lower quality.
Interesting, though as always I think we’re in a grey territory when we try to come up with supposedly objective assessments of the quality of ideas. Nevertheless, I think many of the “builds” in traditional brainstorming are a bit half-hearted so the result is not too surprising.
And this research fits neatly with the link I posted earlier on the value of solitude.