So it turns out that research shows brainstorming doesn’t work.

The reason brainstorming has intuitive appeal is that it encourages you to lay off the censoring for a while, and to go for quantity over quality. Those points are still valid. But the technique needs refinement to be more effective and to yield ideas of a higher quality.

One study, by Gallupe and Cooper (1993) found that electronically mediated brainstormers generated more high quality ideas than face-to-face brainstomers. So to have better face-to-face brainstorming sessions, you should tweak the technique using these guidelines:

  • People should be encouraged to list ideas before coming to brainstorming sessions.
  • The number of ideas produced by each person should be monitored.
  • Problems should be broken down and group members should brainstorm components.
  • Groups should take breaks from each other.
  • High standards should be set for the number of ideas.

It seems that the real benefit of collaborating and working in groups is not in producing ideas, but evaluating them. So a good method for a writing team would be to generate ideas individually first, before meeting up to bandy the ideas around, discuss them, and see which ones sink or swim. New ideas will emerge in these sessions, building on the already existing ideas.