Table 3.

Statistical analysis of the classification accuracy. In this table, ‘n’ is the number of trials; ‘z-scores’ indicates how many standard deviations a particular accuracy level is better than that expected by chance.

	Overt Speech		Covert Speech
	Vowels (n)/(z-scores)	Consonant (n)/(z-scores)	Vowel (n)/(z-scores)	Consonant (n)/(z-scores)
A	72/3.8	72/5.5	72/4.1	72/5.9
B	108/3.6	108/2.03	108/1.2	108/2.08
C	216/5.4	216/4.8	180/2.8	180/1.3
D	216/3.9	216/7.7	72/3.2	72/2.3
E	216/4.7	216/5.4	252/2.48	252/1.44
F	72/3.2	72/2.1	72/2.7	72/1.91
G	144/5.3	144/5.7
H	72/2.66	72/1.3
p-value	p<0.004	p<0.022 (except subject H)	p<0.007 (except subject B)	p<0.03 (except subjects C and E)