Modeling the beneficial effects of incubation in creative brainstorming

A series of recent neural network models have begun to address experimental findings showing that group brainstormers are often less productive than the same number of individuals working in isolation. These models also deal with various manipulations that can enhance productivity in both group and individual brainstormers. The present study uses network interactions between nodes representing ideas and categories to reproduce some of the behavioral data about interactive groups instructed to brainstorm on an open-ended real-life problem. In particular, we simulate results showing that brief breaks or incubation periods can enhance productivity in brainstormers who would otherwise be fixated on categories in which they receive suggestions. Future refinements of this model may be used to better predict group participants' search of ideas.