Figure 1.
Two representative exemplars of the 9 male voices and 9 female voices were illustrated. The consonant part (/h-/) of the sound was normalized at around 40 ms in duration and the whole sound was normalized at 300 ms in duration.
Figure 2.
In Experiment 1, the 9 male voices and 9 female voices were pseudo-randomly presented to participants with equal probability. A pure tone sound was presented after each voice. Participants were instructed to press the female or male key when hearing the pure tone to indicate the gender information of the voice just before the pure tone. In Experiment 2, the 9 male voices and 9 female voices, and a pure tone stimulus were pseudo-randomly presented to participants. Participants were instructed to press the key when hearing the pure tone stimulus.
Figure 3.
Grand-averaged ERP to opposite-sex voices and that to same-sex voices at electrode site PO3.
In Experiment 1, the ERP to opposite-sex voices is positively deflected than that to same-sex voices at parieto-occipital recording sites (cluster p value = 0.014). In Experiment 2, there is no significant difference between the ERP to same-sex voices and that to opposite-sex voices. Gray bars represent the time-window (720–760 ms) centered on the peak of the difference between the ERP to same-sex voices and that to opposite-sex voices in Experiment 1. The topographies illustrate the difference between the ERP to same-sex voices and that to opposite-sex voices at the time-window (720–760 ms).