Change scores in cognitive functions measures of both groups.

Change scores were calculated by subtracting the pre-cognitive measure score from the post-cognitive measure score. We conducted the two types of permutation tests of an analysis of covariance (ANCOVA) for the change scores in each of cognitive tests. In the first ANCOVA, pre-training scores in each cognitive test, sex, and age were the covariates. The results were presented in lines of results of ANCOVAs. In the second ANCOVA, pre-training scores in each cognitive test, sex, age and pre-training score of SRT were the covariates. The results were presented in lines of results of additional ANCOVAs. The level of significance was set at pMoreover, we report eta square (η2) as an index of effect size. As a descriptive index of strength of association between an experimental factor (main effect or interaction effect) and a dependent variable, η2 is defined as the proportion of total variation attributable to the factor, and it ranges in value from 0 to 1. η2≥0.01 is regarded as a small effect, η2≥0.06 as a medium effect, and η2≥0.14 as a large effect.SD, standard deviation. Fluid intelligence was measured using Raven’s Advanced Progressive Matrices Test (RAPMT). Executive functions were measured using Wisconsin Card Sorting Test (WCST) and Stroop Task (ST). Working memory was measured using the Operation Span (OpS), letter–number sequence (LNS), arithmetic (Ari). Short-term memory was measured using the Digit Span (DS) and Spatial Span (SpS). Attention was measured using the Digit Cancellation Task (D-CAT) and Simple Reaction Time (SRT). Processing speed was measured using the Digit Symbol Coding (Cd) and Symbol Search. Visuo-spatial ability was measured using the Mental Rotation task (MT). Reading (verbal) ability was measured using the Japanese Reading Test (JART).