Fig. 2

Exploratory analyses. Panel a Overview of accuracy scores separated by stimulation intensity for both sub-tasks. Data are the same as in Fig. 1a, except that the active stimulation group has been split into different stimulation intensity categories (moderate stimulation (< 26.25), and high stimulation (> 26.25) groups, sham = 10), based on a median split. Error bars reflect standard error of the mean. Note: median split has been conducted for visualization purposes only. Statistical analyses were done using stimulation intensity as continuous variable. Panel b Violin plot of the mean stimulation intensity across four days in both the McIntire et al., [2] and current study. Panel c Relationship between stimulation intensity and accuracy for both sub-tasks per day. The median is indicated with the vertical blue dotted line. Panel d The proposed inverted-U dose-response curve: moderate stimulation intensities are more optimal for improving learning performance compared to too low or too high stimulation.