Table 1.

Classification analysis statistics for visually cued, aurally cued, and memorized trial type experiments.

	Visual cues	Aural cues	Memorized trial type
Number of subjects	6	7	7
Mean number of trials per subject	476	197	538
Mean peak amplitude-based classification rate	0.819	0.914	0.531
95% confidence interval	0.786–0.850	0.873–0.943	0.495–0.567
p-value	0.021*	0.025*	0.40
GC classification rate	0.792	0.605	0.580
95% confidence interval	0.756–0.824	0.545–0.663	0.544–0.616
p-value	3.4 × 10−9*	0.0026*	0.0056*

Mean number of trials indicate the combined number of free and forced trials for all folds. The p-values for Fisher discrimination are calculated at the peak mean classification rate and compared to a chance rate of 0.5 using t-tests. Asterisks here denote significance when controlling for a false discovery rate of α = 0.05 across classification accuracies calculated for every investigated point in time. The p-values for GC classification are given for final validation classification rates compared to a chance rate of 0.5 using t-tests with asterisks denoting significance. Classification rate confidence intervals are calculated for counts aggregated across subjects and by using the Clopper-Pearson method for binomial distributions.