Encoding processes and the spacing effect.

The differential encoding theory of the spacing effect was tested utilizing Martin's (1968) encoding variability notion, in which it is hypothesized that low-meaningfulness items are more variable in their encodings than are high-meaningfulness items. In a series of three experiments using a continuous paired associate learning task, it was predicted that the spacing vs. performance curves for CCC items would show a faster improvement in performance than would the curves for high-meaningfulness CVC items. None of the experiments supported this prediction. In addition, it was found that items recognized on their second presentation were more likely to be recalled than were those items not recognized. It was concluded that an item's repetitions are more effective if one code is formed and elaborated with each repetition rather than if more than one code is formed.