Can working memory training improve working memory capacity in typically developing children?: Training methods reveal specific effects.

This study aimed to investigate whether different working memory training (WMT) tasks differ in their effects on improving working memory, as well as their underlying cognitive neural mechanisms, from the perspective of training task characteristics. In the present study, we compared two tasks: adaptive running memory span (RMS) training and working memory animal span (WMAS) training. Eighty-four children were randomly assigned to one of the two training groups or the control group. All training protocols consisted of 20 sessions over approximately 4 weeks. Before and after training, the children completed a battery of cognitive tasks to examine the transfer of training effects to untrained measures: the backward digit span, which measured phonological WM storage; the change-detection task, which measured visual WM storage; and the n-back task with electroencephalogram (EEG), which assessed multi-item WM capacity. Compared with the control group, the two training groups showed distinct improvements in working memory capacity, which varied by training task: specifically, children in the RMS group exhibited gains in visual WM storage, updating ability, the mean amplitudes of P2 and P3, and theta power. In contrast, those in the WMAS group showed changes in the mean amplitude of P2 and alpha power. These findings provide evidence that WMT may enhance children's working memory capacity; however, such near-transfer effects are modulated by the specific training task.