The offline production effect.

People remember words they say aloud better than ones they do not, a result called the production effect. The standing explanation for the production effect is that producing a word renders it distinctive in memory and thus memorable at test. Whereas it is now clear that motoric production benefits remembering over nonproduction, and that more intense motoric production benefits remembering to a greater extent than less intense motoric production, there has been no comparison of the memorial benefit conferred by motoric versus imagined production. One reason for the gap is that the standard production-by-vocalization procedure confounds the analysis. To make the comparison, we used a production-by-typing procedure and tested memory for words that people typed, imagined typing, and did not type. Whereas participants remembered the words that they typed and imagined typing better than words that they did not, they remembered the words they typed better than the ones they imagined typing; an advantage that was consistent over tests of recognition memory and source discrimination. We conclude that motoric production is a sufficient and facilitative (but not a necessary) condition to observe the production effect. We explain our results by a sensory feedback account of the production effect and sketch a computational framework to implement that approach.