Figure 3.

Properties of the Predictive Model

(A) Spatial support. Upper panel shows mean detection thresholds of three observers for in-phase (black) and antiphase (green) targets, plotted as a function of inducer length. For comparison, the open symbol shows performance with no inducer stimuli. The growth of the interference effect is summarized in the lower panel, which shows the difference between in-phase and antiphase thresholds along with the best-fitting exponential function (semisaturation space constant = 15 arcmin).

(B) Spatial projection. Thresholds are shown as a function of the size of the spatial gap introduced between target and inducer stimuli. The interference effect is restricted to a small spatial region ahead of the inducer (space constant of exponential fit = 25 arcmin).

(C) Time course. Thresholds plotted as a function of the duration of target and inducer stimuli reveal the buildup of the interference effect over time (temporal semisaturation constant = 122 ms).

(D) Interocular transfer. Comparison of performance in conditions where target and inducer stimuli were presented either to the same eye (monocular) or to different eyes (dichoptic) suggests that predictions formed based on motion information presented to one eye can interfere with visual input to the other.

Error bars indicate ± 1 standard error.