Abstract
Introduction: Ensemble perception is attracting increasing interest. Presented with element arrays, participants accurately report means of different array features, but are unable to report its members’ features. Ensemble perception studies used spatial presentation, i.e. simultaneously presenting elements scattered on the screen. Recently, Khayat and colleagues (2018, 2019, 2020) presented temporal sequences of ensemble elements, studying implicit effects of ensemble mean and range. We now use temporal and spatio-temporal presentation, explicitly asking participants to match a test element to sequence mean, or judge presence or absence from the sequence. Methods: We presented a temporal element sequence, all at fixation or randomly scattered, asking MTurk participants to relate to circle size, line orientation, or disc brightness. Following the sequence, participants saw a single test element, and reported presence/absence of that size/orientation/brightness in the sequence (membership test), or if the test element was larger/more clockwise/brighter than the sequence mean (mean evaluation test). Results: Participants judged test elements as present in the sequence when they were at or near sequence mean, and absent when far from the mean, or outside the sequence range – for both temporal and spatio-temporal sequence presentation. They clearly perceived sequence mean and range for size/orientation/brightness features. Nevertheless, they did not differentiate between test elements that were/were not in the sequence, except for rejecting out-of-range elements. Participants were very accurate in the mean evaluation test, with accuracy increasing (near the mean) by about 5% per 3% feature change. Interestingly, there was little difference between spatio-temporal and pure temporal sequence presentation at fixation. With our parameters, evaluation of circle size was less uncertain than other features, and brightness most difficult. Conclusions: Participants are excellent at explicitly judging sequence mean and range, for size, orientation and brightness features; are blind to presence/absence of specific elements; irrespective of foveal versus scattered presentation.