Skip to main content
. 2022 Mar 29;9:114. doi: 10.1038/s41597-022-01226-4

Fig. 9.

Fig. 9

Classification accuracy from a positive control MVPA “buttons” analysis (classifying whether examples were after a button-press or not), Sternberg task. Leave-one-subject-out cross-validation within surface Schaefer 1000 × 17 parcels81, linear SVM (c = 1), default scaling, chance is 0.5. Classification accuracy was greatest in visual and left somatomotor parcels, as expected since the right hand was used for button pressing and the stimuli screens varied in appearance between response and baseline periods (Fig. 3). Code and additional analyses in the controlAnalysis files17.