Abstract
Our knowledge about relevant objects in our environment reflects an integration of current visual input with information acquired during preceding gaze fixations. For this accumulation of input to be of benefit, the visual system must determine what information should be integrated and what should be segregated. To investigate the basis of this decision, in Experiment 1, participants first viewed a colored disk in their peripheral vision, then made a saccade that shifted the object into the opposite hemifield. During the eye movement, the object underwent changes of varying magnitude in its color and/or location. Participants were asked to detect whether the object had changed in any dimension as well as report the post-saccadic color from a color wheel. Integration was observed as a bias in the color report in the direction of the pre-saccadic color. The magnitude of the color change was found to influence the bias in color reports, indicating reduced integration for larger changes. In contrast, changes in the object’s location, while detected with similar frequency, had no significant effect on the integration of pre-saccadic with post-saccadic color. Experiment 2 replicated this finding using orientation as the secondary feature dimension in place of location. In Experiment 3 participants were unpredictably asked to report either color or orientation on each trial, equating attention. Here, changes in orientation reduced the amount of transsaccadic color integration, but only weakly. These results suggest that the decision to integrate or segregate feature information across fixations depends on whether a change in that feature is detected, but is largely independent of changes in other features of the same object.