Table 4.

Within-subject similarity of topographies based on different measures of spindle activity.

Comparison				Correlation		Classification
			N	R	Pcorr < 0.05	I ≥ II	II ≥ I
Sigma power vs. Spindle density	Slow	N2	24	0.17 ± 0.38*	46%	17%**	13%*
		N3	24	0.62 ± 0.23***	88%	58%***	75%***
	Fast	N2	26	0.24 ± 0.26***	42%	19%***	27%***
		N3	26	0.41 ± 0.23***	73%	54%***	50%***
Sigma power vs. Spindle amplitude	Slow	N2	24	0.16 ± 0.32*	38%	21%***	17%**
		N3	24	0.46 ± 0.26***	71%	13%*	21%***
	Fast	N2	19	0.44 ± 0.23***	77%	42%***	31%***
		N3	18	0.56 ± 0.22***	54%	35%***	35%***
Spindle density vs. Spindle amplitude	Slow	N2	24	0.38 ± 0.37***	71%	25%***	13%*
		N3	19	0.42 ± 0.36***	63%	21%***	29%***
	Fast	N2	26	0.47 ± 0.20***	84%	42%***	35%***
		N3	18	0.27 ± 0.25***	38%	15%**	19%***

R: Pearson correlation coefficient; P_corr < 0.05: percentage of subjects with False Discovery Rate-corrected P values < 0.05; I ≥ II and II ≥ I: classification direction. Significance levels (correlation: one-sample t-test vs. zero; classification: binomial test) indicated by * < 0.05; ** < 0.01; *** < 0.001. Note: cases where spindles were not detected on every channel were excluded from spindle amplitude analyses.