Results from the preliminary experiment. The effect of noise level and WM tone level on presence ratings. The left panel displays the data from the flankers-absent task, and the right panel displays the data from the flankers-present task. Each data point represents the mean rated WM tone presence. Error bars represent ± 1 inter-subject standard error. The data are offset slightly along the abscissa to allow the reader to distinguish clearly between separate data points and error bars. The WM tone was completely silent (i.e., absent) after infinite (∞) attenuation.
Illustration of a condition tested in experiment 1. Each line represents a pure tone and each shaded box represents a low-pass filtered white noise. Both panels display the stimuli in which the WM gap duration was set to 60-ms, and the WM gap occupied a central position. The flanker tones are present in both cases. The interval displayed in the top panel (a) comprises a WM silence and that in the bottom panel (b) comprises a WM tone.
Results from experiment 1. The effect of WM gap duration, position, and flanker condition on discrimination between a WM tone and a WM silence. In the main panel (top), each data point represents the mean percentage of correct responses from all listeners (N = 15). The bottom left panel displays only the data from the four listeners who performed very accurately in the flankers-present conditions. The bottom right panel displays the data from the remaining 11 listeners only. Error bars represent ± 1 inter-subject standard error.
Illustration of a two flanker-tone conditions tested in experiment 2. Both panels display the stimuli comprising a 60-ms WM tone. In the top panel (a), the flanker tones are aligned along a common ascending-frequency trajectory (the aligned-glide condition). In the bottom panel (b), both flanker tones descend in frequency but the onset frequency of the second tone is offset in relation to the first flanker tone (the offset-glide condition).
Results from experiment 2. The effect of WM gap duration and flanker condition on discrimination between a WM tone and a WM silence. Each data point represents the mean percentage of correct responses (N = 15). Error bars represent ± 1 inter-subject standard error.
Results from experiment 3. The effect of flanker type and masker type on rated perceived continuity. Each data point represents the mean rated continuity for each masker type (N = 15). Error bars represent + 1 inter-subject standard error.
Article metrics loading...
Full text loading...