Temporal analysis of image-rivalry suppression

Abstract

During binocular rivalry, perception alternates between two different images presented one to each eye. At any moment, one image is visible, dominant, while the other is invisible, suppressed. Alternations in perception during rivalry could involve competition between eyes, eye-rivalry, or between images, image-rivalry, or both. We measured response criteria, sensitivities, and thresholds to brief contrast increments to one of the rival stimuli in conventional rivalry displays and in a display in which the rival stimuli swapped between the eyes every 333 ms-swap rivalry-that necessarily involves image rivalry. We compared the sensitivity and threshold measures in dominance and suppression to assess the strength of suppression. We found that response criteria are essentially the same during dominance and suppression for the two sorts of rivalry. Critically, we found that swap-rivalry suppression is weak after a swap and strengthens throughout the swap interval. We propose that image rivalry is responsible for weak initial suppression immediately after a swap and that eye rivalry is responsible for the stronger suppression that comes later.

Figure 1. Illustration of stimuli and procedure during flicker-and-swap (FS) rivalry for a trial when a probe was delivered (panel A).

For simplicity, we have shown neither that the stimuli were flickering on and off at 18 Hz nor that the screen went black 100 ms after the presentation of the probe until the observer had responded. Illustration of the time course of the probe (panel B).

Figure 2. Mean strengths of suppression for different rivalry conditions from Experiment 1. Error bars show 1 standard error of the mean.

Figure 3. Response criteria measure for different rivalry condition in Experiment 2 at the common contrast increment of 0.27 (n = 5) for all test conditions.

Error bars show ±1 standard error of the mean. The differences are significant: Criteria are conservative (observers are unlikely to say the probe was presented) for Static rivalry, neutral for FO, and liberal for FS.

Figure 4. Strength of suppression for different rivalry conditions in Experiment 2.

The panels on the left show the strength of suppression for static (top), FO (middle), and FS (bottom) as a function of contrast (n = 3). The panel on the right shows the strength of suppression for these conditions for a contrast increment of 0.27 (n = 5). Error bars show 1 standard error of the mean. There are neither significant differences among contrasts nor among the three rivalry conditions.

Figure 5. Temporal analysis of strength of suppression during swap interval of Experiment 2.

Mean strength of suppression for probes as a function of when they reached maximum contrast after a swap during swap rivalry. Probes began about 50 ms earlier than this and persisted for about 50 ms after this. Probes reaching full contrast 283 ms or more after a swap were increasingly displayed during the early times after the next swap. The contrast increment was 0.27 (n = 5). The curve shows the significant quadratic function that explains some of the variability of the FS data.

Figure 6. Strength of suppression for different rivalry conditions in Experiment 3 as a function of when they reached maximum contrast after a swap during swap rivalry (there was no swap for FO rivalry).

Error bars show ±1 standard error of the mean. There is a significant strengthening of suppression with time in the swap interval.