Selection of information necessary for successful self-derivation.

Accumulation of knowledge relies in part on self-derivation of new semantic knowledge through integration of separate yet related learning episodes. Prior research suggests that individual and developmental variability in self-derivation is due to differences in the precursor processes of encoding, reactivation, and integration. In the present research, we examined a fourth potential precursor process: selection of learning episodes most relevant to the target self-derived knowledge. In two experiments, we examined selection of information most relevant to self-derivation in 8-year-olds (Experiments 1 and 2) and 12-year-olds (Experiment 2). Both age groups self-derived even when there were several candidate facts from which to select. Older children had higher levels of self-derivation performance and made more correct selections than younger children. Within and across age groups, selection performance significantly predicted trial-level self-derivation success. These data provide evidence that selection of facts necessary for self-derivation contributes to the robust variability observed in self-derivation. • The work examines selection of relevant facts and self-derivation in children. • 8- and 12-year-olds self-derive in both high- and low-selection demand conditions. • Children were more likely to self-derive if they selected at least 1 relevant fact. • Selection of relevant facts contributes to age-related and individual differences. [ABSTRACT FROM AUTHOR]