Abstract
Summary statistics (e.g., average) are instantaneously extracted from multiple objects. This is a kind of automatic cognitive processing without much effort. When objects are categorized by a distinct feature (i.e., color), summary statistics of objects with each feature (i.e., red) can be extracted. This indicates a possibility that summary statistics are calculated with attention guidance based on saliency. Attention also has a function of marking objects and excluding them to facilitate processing efficiency even when they have the same physical features (visual marking and preview benefit: Watson & Humphreys, 1997). In this study, I investigated whether volitional attention can organize objects for summary statistics calculation using a preview paradigm. In the whole condition, all dots were simultaneously presented, and the participants clicked the average location of these dots. In the preview condition, a part of dots were initially presented to the participants, and followed by remaining. Participants were asked to indicate the average location of later dots ignoring the initial dots. Experiment 1, in which later dots were simply added to the initial dots, showed that the participants could indicate the average location ignoring the initial dots. That is, error of the average location of dots was not differed between when they were added to the initial display (preview condition) and when they were simultaneously presented (whole condition). Experiment 2, in which a brief blank was inserted before the appearance of the later dots in the preview condition, showed that the preview benefit disappeared. That is, error of the average location of the latter dots in the preview condition was much larger compared with in the whole condition. These results suggest that summary statistics can work with attention guidance based on saliency, but cannot work with volitional attention guidance.