Abstract
Our previous investigations on the human estimation of temporal average from dynamically changing visual stimuli showed that observers often emphasize information at ~200-500 ms before decision to judge the temporal average (Sato et al., 2013, 2016). This recency effect is found for a range of visual attributes including orientation, motion direction, and facial expression. Exceptionally, when observer judge the arithmetic average of digit numbers, they equally weigh information at any temporal location over the stimulus presentation (Sato et al., 2016). In order to examine whether the lack of recency effect in the judgement on temporal average of number is specific to digits, the present study investigated human judgments on temporal average of the number of dots. Stimulus was a texture composed of white dots with maximum quantity of 2-40, and the quantity of dots at each frame was temporally varied in accordance with a Gaussian distribution with a particular mean and S.D. during the presentation of 2 sec. Observers were asked to judge whether the average number of dots was above or below a certain number. By means of a logistic reverse correlation analysis, we calculated the impact of information at each temporal frame upon the observer's response. The results revealed no recency effect again for the temporal average of dot number. This seems to indicate that numerical average over time is estimated by a common mechanism regardless of whether information was given as digits or dot textures, which is distinct from mechanisms involved in estimation of temporal average of the other visual attributes.
Meeting abstract presented at VSS 2017