Figure 2.

The ventriloquism after-effect. Regression of mean auditory localization judgments taken before (pre-training, x-axis) and after (post-training, y-axis) exposing human subjects to spatially misaligned auditory and visual stimuli. (a) Data obtained after conditioning with an 8° mismatch between the stimuli. The dashed line indicates perfect correlation between the pre- and post-training estimates. In this case, the auditory localization estimates have been shifted in the direction of the previously present visual stimulus, so the regression line lies above the dashed line. (b) Data obtained after conditioning with a 0° mismatch between the stimuli. There is no difference between pre- and post-training estimates and the dashed line is hidden by the thicker solid regression line. Adapted and reproduced with permission from Recanzone (1998). Copyright (1998) National Academy of Sciences, USA. Open and filled squares, subject 1; open and filled diamonds, subject 2; open and filled circles, subject 3 (open symbols, 750 Hz; filled symbols, 3 kHz).