III. The Experiment

Fortunately, it is possible to make the same sort of changes without monitoring eye movements and making changes during a saccade.Ron Rensink (1997) has developed a "flicker paradigm" to make changes in complex scenes. We will use his technique here to test the hypothesis that if we have an integrative visual buffer, then we should easily detect changes that occur in a complex scene. Here's the basic idea of the experiment:

Figure 1. General design. A and A' are two versions of a picture that differ from each other. The experiment begins by displaying picture A. After a short period, a blank field is presented, then picture A' is presented. If we have an integrative visual buffer that is representing these two scenes as they are presented over time, as detail is acquired from both scenes, any discrepancy will quickly and readily become apparent . . . or will it?