Table 4.

Multinomial logistic model output for Experiment 3, indicating the likelihood of naïve listeners selecting the voice modulation expressing the intended social trait related to a scenario, relative to voice modulations expressing alternative and less relevant social traits

Scenario-related intended trait	Alternative choice	OR	β	SE	z	p
Hostile	Confident	3.12	1.14	0.06	18.24	<0.001
	Likeable	5.21	1.65	0.08	21.55	<0.001
Likeable	Confident	1.53	0.42	0.06	7.70	<0.001
	Hostile	3.75	1.32	0.08	17.51	<0.001
Confident	Hostile	2.02	0.70	0.07	10.53	<0.001
	Likeable	1.15	0.14	0.06	2.50	<0.05

OR = odds ratio; SE = standard error