Table 1.

Linear Mixed-effects Model of the Response Amplitude

Fixed Effects	Estimate	SE	df	t	p
Baseline	183.1	24.8	33.7	7.4	10−8
Auditory noise (An)	−109.6	13.5	188	−8.1	10−13
Posterior location × An	140.6	21.2	188	6.6	10−16
Posterior location	−101	38.7	34.2	−2.6	.01
Visual noise (Vn)	21.6	13.5	188	1.6	.11
An × Vn	−13.3	19.1	188	−0.7	.49
Posterior location × Vn	−8.9	21.2	188	−0.4	.67
Posterior location × An × Vn	3.6	29.9	188	0.1	.91

Results of an LME model of the response amplitude. The fixed effects were the location of each electrode (Anterior vs. Posterior), the presence or absence of auditory noise (An) in the stimulus and the presence or absence of visual noise (Vn) in the stimulus. Electrodes and stimulus exemplar were included in the model as random factors. For each effect, the model estimates (in units of percent signal change) for that factor are shown relative to baseline, the response in anterior electrodes to clear audiovisual speech (AV stimulus condition). The “SE” column shows the standard error of the estimate. The degrees of freedom (“df”), t value, and p value derived from the model were calculated according to the Satterthwaite approximation, as provided by the lmerTest package (Kuznetsova, Brockhoff, & Christensen, 2015). The baseline is shown first; all other effects are ranked by absolute t value. Significant effects are shown in bold. The significance of the baseline fixed effect is grayed-out because it was prespecified: only electrodes with significant amplitudes were included in the analysis.