Incubation benefits only after people have been misdirected

Incubation refers to the popular idea that stopping work on a problem may, at times, be a more efficient means by which to reach a solution than continuing to work. Empirical studies of incubation have used few participants and have provided ambiguous and discrepant results. We investigated three potential accounts of incubation in retrieval and search problems (subconscious work, spreading activation, and fixation forgetting) with the help of a large Internet-based participant pool. The amount of time allotted for explicit work on each of 12 problems was controlled, while the distribution of that time was manipulated in several incubation conditions. When problems were presented by themselves, none of the incubation conditions aided in the solution of either anagram or remote associate test problems. However, incubation benefits arose when participants were given misdirecting clues (probably because time delays facilitated forgetting of these clues).