Procedure

Observers were first presented with an animation of the target and context figure, or without the context during baseline trials, moving across the display. Then, the target, and context if applicable, disappeared, and the display was rotated by changing the camera’s perspective to present the goal posts in frontal view; see Figure 2 for an example trial. Observers were instructed to indicate via a mouse-click the location where they thought the target would cross the goal.

Observers were presented with one of three possible target sphere start and end locations. The start locations are described by the following XYZ Blender positions. See Figure 3 for an illustration of each of the possible starting locations.

Location 1: starting location (–7.2, –1.5, 1.33), ending location (7.2, 2.5, 1.45)

Location 2: start location (–7.2, –3.5, 1.33), ending location (7.2, 0.5, 1.45)

Location 3: start location (–7.2, –5.5, 1.33), ending location (7.2, –1.5, 1.45)

Regardless of the start or end location of the target sphere, its motion followed a linear trajectory oriented at 4° from horizontal.

Each target-sphere animation was accompanied by one of four context (secondary sphere or cast shadow) trajectories, as well as a no-context baseline trajectory. For all context trajectories, the XY Blender start locations were identical to those of the target-sphere, with Z Blender space set to a value 0.41, and the context’s X Blender location adjusted to be perceptually below the target. The following context trajectories, illustrated in Figure 3, were tested:

Follow: The Y Blender position was the same as the top-sphere and followed the same linear trajectory oriented 4° from horizontal.

Intermediate: A value of 2 units was subtracted from the Y Blender ending position of the top-sphere. This motion trajectory followed a linear path orientated 2° from horizontal.

Horizontal: A value of 4 units was subtracted from the Y Blender ending position of the top-sphere. This motion trajectory followed a horizontal path, that is, 0° from horizontal.

Down: A value of 6 units was subtracted from the Y Blender ending position of the top-sphere. This motion trajectory followed a linear path oriented –2° from horizontal.

Horizontally flipped versions of all conditions were also presented. All conditions, horizontally flipped and nonflipped, were interleaved and tested 10 times.