Sequential dependency in percept durations for binocular rivalry

Abstract

When the two eyes are presented with different images, perception cycles through a sequence of dominance durations for either image: binocular rivalry. Generally, the dominance durations are assumed to be sequentially independent. However, a moderate but consistently positive sequential dependency of up to 20% has been reported across studies that took the step to analyze sequential dependency. This dependency is not only consistently present but also consistently ignored for interpretation. Here we investigated this sequential dependency in dominance durations both experimentally and theoretically. We presented orthogonal gratings and subjects reported the experienced percept (taking the durations of mixed percepts into account). We found a positive sequential duration dependency consistent with existing literature, underscoring the reproducibility of this effect. Furthermore, we found, by varying the contrast of the gratings across trials, that this sequential dependency increased with increasing contrast of the gratings. Statistical data simulations showed that this correlation between stimulus content and sequential dependency do not naturally result from the decrease in average percept durations with increasing contrast. It is clear that the sequential dependency and its correlation with stimulus content provide important information on the mechanism underlying binocular rivalry. We found that our results are an emergent property of a neural model by our group (Noest et al, VSS 2006), originally designed to explain perceptual memory for interrupted ambiguous stimuli.