Figure 2.

Offline selection accuracy of the healthy participants. Accuracies are shown separately for vowels, consonants, and Hiragana syllables (the latter being correct if both the correct vowel and consonant were selected). The accuracies do not take the decrease of selection times into account that occurred due to the reduction of stimulus repetitions. The second row shows the information transfer rates (ITRs) for vowel, consonant, and Hiragana selection (which conveys the most information per selection) of the healthy participants. The ITRs take the decrease of selection times into account that occurred due to the reduction of stimulus repetitions. The third row the time per selection needed by the healthy participants after the number of stimulus repetitions had been adjusted based on the performance in the calibration session.