| In real world,the amazingly efficient integration of massive information is operated by limited cognitive resources.To balance speed with accuracy in such integration,the spatio-temporal integrated statistical summary representations(SSRs)play an important role.Yet,the processing characteristics and mechanism of such representations are still unclear,especially when dealing with hierarchical structure ensembles.This issue was investigated in the present study.Experiment 1(a and b)explored the characteristics and mechanism of the spatial integration of SSRs,while experiment 2(a and b)explored the temporal integration.Same set of 18 discs with various sizes was used in all experiences in this study.These discs were divided into two groups with different set sizes and presented either simultaneously(experiment 1)or sequentially(experiment 2).Observers were asked to view these two groups of stimuli and report the mean size of either group or the whole.All experiments showed the same pattern that the reported mean sizes of each group were deviated from true values and bias to the whole.However,the reported mean sizes of the whole were the same between experiment 1a and 1b,but significant different between experiment 2a and 2b,which actually have the same value in math.Moreover,in experiment 2a and 2b,the reported mean size of the whole could be predicted by the pooled mean calculated based on the reported means of each groups.Thus,the main findings of this study are as follow:The integration of SSRs of multiple ensembles is a hierarchical process with partly different patterns in space and time.The spatial integration of SSRs constructs different levels of SSRs independently,while the temporal integration constructs the higher level SSRs based on the lower level SSRs.More importantly,visual system gives quite precise weight to each part in the whole according to their numbers when constructing the higher level ensemble in time period. |
PDF Full Text Request