Abstract
We tested the relative contributions of local interactions and position-independent analyses to object substitution masking by embedded the target and mask in an attentive tracking task. Six discs were presented circling the central fixation at a constant radius and speed; one disc was identified as the target to be tracked. Tracking continued for an average of 2 seconds and on 10% of trials, tracking accuracy was evaluated with a probe display. The remaining trials ended with a masking display. All the discs changed briefly to Landolt C and one of them was surrounded by four small dots. All items remained in motion during the masking phase. The dots indicated which item to report and also served as the mask of the probed item. To assess mask integration time, a blank ISI of variable duration (0 to 450 ms) was inserted between the initial 80–120 ms (varied across subjects) phase when both the mask and target were presented together and the 80 ms trailing phase when only the mask was present. Classic object substitution masking was seen at the short ISIs and neither the masking nor integration time was affected by the rotation speed during the masking phase (although a static presentation prior to the test generated an afterimage that degraded performance). Our results suggested that object-based, position independent masking was more important than any local interactions.
Supported by JSPS for TY.