Fig. 4. Motor skill acquisition in experiment 2.

(A) Average scores generated during experiment 2, consisting of motor training with stimulation, with the error bars depicting the SEM. The data are grouped by age group and stimulation type [i.e., verum (V) or placebo (P)]. (B) Percentage of total learning over the entire training week represented by fast online learning (D1), online learning during D2 to D5, and offline learning between training days. Each outer ring captures the proportion of total learning by these three aspects, while the inner rings present their time course during the week (counterclockwise): The first inner circle is the online performance gain during D1, the second inner circle is the offline performance gain between D1 and D2, the third inner circle is the online gain during D2, etc. Yellow and blue represent improvements, while black and gray represent worsening of performance. Please note the large difference in regard to the proportion of total learning explained by fast online learning between the verum and placebo groups in older participants. (C) Speed and accuracy normalized to the first block of training, grouped by age group and stimulation type. Please note that while most groups show similar dynamics to those seen in the respective age-matched groups from the first experiment, the older group receiving verum stimulation shows dynamics more similar to those seen in young adults. As in the first experiment, the accuracy was maintained even at increasing speeds, although young adults receiving verum stimulation significantly dropped in accuracy on the last training day. The shading represents the 95% confidence interval for the logarithmic curve fitting (the type of curve is for display purposes only and not included in the LME analysis).