Table 3.

Detailed test statistics for the three-way ANOVA on mean percent correct and response times as dependent measures (Analysis 1 in the main text).

	Percent correct				Response times
Effect	F (dfn,dfm)	p	${\mathrm{\eta }}_{\mathrm{p}}^{\mathrm{2}}$	𝜀	F(dfn,dfm)	p	${\mathrm{\eta }}_{\mathrm{p}}^{\mathrm{2}}$	𝜀
Set size	199.01 (1,47)	<0.001	0.81		16.55 (1,47)	<0.001	0.26
Cue type	61.77 (2,94)	<0.001	0.57	0.88	146.84 (2,94)	<0.001	0.76	0.72
Second delay	0.35 (3,141)	0.787	0.01		0.58 (3,141)	0.596	0.01	0.82
Cue type × set size	0.89 (2,94)	0.413	0.02		1.97 (2,94)	0.155	0.04	0.79
Cue type × second delay	11.69 (6,282)	<0.001	0.20	0.78	15.53 (6,282)	<0.001	0.25	0.72
Set size × second delay	2.57 (3,141)	0.056	0.05		0.50 (3,141)	0.656	0.01	0.85
Set size × cue type × second delay	0.28 (6,282)	0.945	0.01		1.43 (6,282)	0.226	0.03	0.68

df_n and df_m denote the numerators’ and denominators’ degrees of freedom for the respective F-test. 𝜀 is the Greenhouse–Geisser estimation of sphericity violations.