Figure 1.
Schematic illustrating the four experimental conditions used in our protocol.
Auditory voice feedback either remained present throughout the trial (A), or was replaced with masking noise after 10 s (NA). Similarly, visual feedback of vocal intensity relative to the 80 dB SPL target either remained present throughout the trial (V), or was removed after 10 s (NV). Following feedback manipulations, participants were instructed to make continued responses synchronized with the visual metronome and at the target vocal intensity for the remainder of the trial. We hypothesized that provision of visual reference stimuli would calibrate attenuated somatosensory signals and result in reduced expression of the Lombard effect following removal of auditory voice feedback.
Figure 2.
The vocal intensity time series for all conditions, grand averaged across participants.
Error bars represent SD. The vertical dashed line represents the time at which auditory voice feedback, visual feedback of vocal intensity or both were removed.
Figure 3.
Significant interactions for overall vocal intensity (A) and for the difference in mean vocal intensity over the two trial phases corresponding to the period before feedback removal (t = 2–10 s) and the period following it (t = 11–18 s) (B) among factors of auditory feedback condition and visual feedback condition.
In all cases, asterisks and connecting lines represent reliable pairwise comparisons, significant at p<.01. Error bars represent standard error.