Figure 1.
Examples of tone patterns for the three experimental conditions.
Patterns were constructed individually from four, six or eight different tones, respectively, and the size and direction of intervals was controlled for between subjects. Time and pitch are indicated on arbitrary scales.
Figure 2.
Example of three standard sequences and one deviant sequence as they might have occurred during the experiment.
In deviant sequences one note at a random position (large dot) was shifted one tone up or down, unless this would be the same pitch as a neighbouring tone. Time and pitch are indicated on arbitrary scales.
Figure 3.
Group data (source waveforms) for the three experimental conditions (four, six and eight tone patterns) and the control condition and for both hemispheres hemisphere in the two groups (musicians = black traces, nonmusicians = grey traces).
Arrowheads indicate the latency of the mismatch responses in the group averages.
Figure 4.
Mean MMN amplitudes for each condition and hemisphere for all subjects.
The analysis of variance indicates a left-lateralized dominance for the six tone condition. Error bars represent standard deviations.
Figure 5.
Lower boundaries (black traces) of generated bootstrapped confidence intervals for signal amplitudes (grey traces) for both groups in the three experimental conditions (four, six and eight tones) and the control condition for both hemispheres.
Lower confidence intervals larger than zero indicate that there is significant deflection of source waveform. This is the case in all experimental conditions in both hemispheres for musicians, but only in the control condition, and in the four tone condition (right hemisphere) for non-musicians.
Table 1.
dPrime averages and standard deviations (in braces) for each condition and group.