Fig. 5. Illustration of response distributions and hypothesis test. A, Predicted distributions of direction estimates under the vector-averaging hypothesis (left panel) and the winner-take-all hypothesis (right panel). The gray panel represents the display monitor; the black circle represents the target ring; magenta points represent individual saccade endpoints; converging dashed lines represent the monkey's direction of gaze. The black downward arrowindicates the direction of visual motion, and the cyan rightward arrow indicates the microstimulated direction.B–E, Distributions of individual direction estimates for four of the visual directions (0, 80, 100, 180°) from Figure 3(ra051b). R values are given where relevant.Black points represent direction estimates for the visual stimulus without microstimulation (visual stimulus coherence was 80%); magenta points represent direction estimates for trials when microstimulation was applied in combination with high coherence visual stimulus; cyan points represent direction estimates on trials when microstimulation was applied during a 0% coherent stimulus. Dashed radius represents the direction of visual motion. Solid black and cyan radiirepresent the median direction estimates for the visual and 0% coherence trials, respectively. The median direction estimate for 0% coherence trials was 17° in this experiment, and the median direction estimates for visual trials, from top to bottom, were 2, 62, 85, and 193°. F–I, Distributions of direction estimates for four visual directions from an additional experiment in another monkey (gr026j). R values are given where relevant. Median direction estimate for 0% coherence trials was 263° in this experiment, and the median direction estimates on visual trials, fromtop to bottom, were 248, 152, 149, and 75°. Visual stimulus coherence was 100%.