Figure 1.

Panel A shows a schematic representation of experimental trials. In different blocks, subjects compared the locations of dot targets either 1) within or 2) between images,when presented against a white noise background, in the absence of any scene. In each trial, dots were presented simultaneously during the first 100 ms, and the background remained otherwise constant throughout the trial (i.e. 1 s). In separate blocks, subjects performed a simple dot detection control task. For each subject, the response accuracy converged towards 75%. Panel B shows a schematic representation of stair case method used to control subjects response accuracy. Since performance on the spatial comparison task varied between subjects, we adjusted the “overall load” by varying the target dot contrast. Thus, for those subjects that had more difficulty in the spatial comparison (i.e. higher spatial comparison demand), dot contrast was increased to ease dot detection. Conversely, for those subjects performed the spatial comparison task more easily (i.e. lower spatial comparison demand), we reduced the dot contrast to make dot detection harder. According to this paradigm, dot contrast varied positively correlated to spatial comparison demand.