Table 1 HbT response mean values averaged across the 21 subjects within the time window from 2s to 18s and p-values indicating statistical significance of the difference between the response and baseline, evaluated using two-tailed Student’s t test for cluster C1 which shows a statistically significant negative mean speech response averaged across the conditions, for cluster C2 which shows a significant negative response to neutral speech.

From: Hemodynamic responses to emotional speech in two-month-old infants imaged using diffuse optical tomography

 

HbT 2–18 s

p uncorr/corr. Two-tailed t test

GM voxels with HbT > 0 for at least one stimulus condition averaged in the FOV (Figs 1 and 2b) Nvox = 28191

 

NMC = 1 ANOVA: p = 0.011** post hoc: happy > neutral**

Speech Average

−0.3

0.39

Neutral

−1.0

6.1 × 10−4

Happy

0.5

0.21

Angry

−0.6

0.055

Sad

−0.3

0.48

C1 Temporal cortex (Fig. 3a–c) Speech average < baseline Nvox = 792 (L3)

 

NMC = 120 ANOVA p = 0.64

Speech Average

−2.5**

3.3 × 10−4/0.04**

Neutral

−3.7*

0.037*/−

Happy

−0.8

0.65/−

Angry

−3

0.084/−

Sad

−2.6

0.13/−

C2 Temporal cortex (Fig. 3d–f) Neutral < baseline Nvox = 471 (L2)

 

NMC = 120 ANOVA p = 0.020* Angry > neutral*

Speech Average

−1.5*

5.3 × 10−3*

Neutral

−4.1**

7.0 × 10−6/8.4 × 10−4**

Happy

−1.2

0.32/−

Angry

−0.2

0.82/−

Sad

−0.6

0.50/−

  1. The p-values for the clusters are given both as uncorrected values as well as corrected for multiple comparisons using Bonferroni (N = 120). For clusters C1-C2, the voxel-wise significance level that gives the optimal cluster-wise significance is stated (L1 pth = 0.001; L2 pth = 0.0033; L3 pth = 0.01) along with the cluster extent. *p < 0.05 uncorrected; **p < 0.05 corrected for the number of regions using Bonferroni method (N = 1 for the global test, N = 120 for clusters).