Table 3.

Reproducibility statistics for the four factorial models.

	Reconstruction Repetitions×Streamline Selection Variant
	Model 1: Within-Session Reproducibility				Model 2: Between-Session Reproducibility
Factor	F	df	p	η2	F	df	p	η2
Repetitions	2268.52	1, 821	<.0001	.0863	442.35	1, 828	<.0001	.0445
Selection Variant	2822.28	2.30, 1884.41	<.0001	.6548	901.92	2.46, 2039.63	<.0001	.4157
Repetitions*Selection Variant	49.49	2.86, 2346.96	<.0001	.0021	50.96	2.67, 2218.82	<.0001	.0043

	Type of Head-Coil × Streamline Selection Variant
	Model 3: Within-Session Reproducibility				Model 4: Between-Session Reproducibility
Factor	F	df	p	η2	F	df	p	η2
Coil	5.88	1, 573	.0153	.0004	110.67	1, 577	< .0001	.0357
Selection Variant	1885.70	2.29, 1310.40	<.0001	.7038	464.87	2.46, 1418.41	< .0001	.2931
Coil* Selection Variant	2.24	2.84, 1625.78	.084	.0002	14.34	2.81, 1621.08	< .0001	.0029

Table gives the ANOVA-type statistics of the nparLD factorial models (see Section 2 and [6] for more details). Repetitions, reconstruction repetitions [1 vs 10 repetitions]; Selection Variant, streamline selection variant [endpoint_nofuzzy, endpoint_fuzzy, visiting_nofuzzy, visiting_fuzzy]; Coil, type of head-coil [12ch vs 32ch]; df, numerator and denominator degrees of freedom (separated by comma); η², eta squared, denoting the share of total variance explained by each factor. As per convention by Cohen [14], an effect is considered small if η²≥.01, medium if η²≥.06, and large if η²≥.14.