Table 1.

Power analyses and exact p values

ReferencetestSample sizeObserved power (α = 0.05)p value
aTextANOVAa.m: 53
p.m: 52
0.150.65
b Fig. 4CKWPre: 347
Post: 311
Back: 340
1.03.3e-78
MW USee previousPre vs post: 1.0
Pre vs back: 1.0
Post vs back: 1.0
Pre vs post: 4.2e-08
Pre vs back: 1.8e-81
Post vs back: 4.9e-29
c Fig. 4E, leftKWSee previous1.02.5e-34
MW USee previousPre vs post: 0.75
Pre vs back: 1.0
Post vs back: 1.0
Pre vs post: 8.9e-04
Pre vs back: 2.1e-38
Post vs back: 3.3e-11
d Fig. 4E, rightKWSee previous1.01.6e-40
MW USee previousPre vs post: 1.0
Pre vs back: 1.0
Post vs back: 1.0
Pre vs post: 7.1e-08
Pre vs back: 1.2e-44
Post vs back: 3.3e-10
eTextWSR3401.06.1e-43
f Fig. 5Ct-test180.577.8e-03
g Fig. 5Ft-test180.683.1e-03
h Fig. 6Bt-test170.982.8e-05
i Fig. 6Et-test170.541.0e-02
jExtended Data Fig. 6-1Bχ226311.01
  • Lower case letters in first column are used as superscripts in text and figure captions. KW, Kruskal–Wallis test; MW U, Mann–Whitney U test; WSR, Wilcoxon signed-rank test; t test, one-sampled t test against zero-mean distribution; p values in rows b–d were corrected using the Bonferroni method.