Figure 2. Mean responses for the (a) unpleasant-pleasant, and (b) rough-smooth response scales for the three overall attenuation levels (0 dB, –20 dB, or –40 dB) against the three frequency manipulations (high frequencies attenuated, veridical auditory feedback, or high frequencies amplified). Each data point reflects the average of 100 trials (5 trials for each participant). Error bars represent the between-participants standard errors of the mean.