Multiple concurrent centroid judgments imply multiple within-group salience maps

Subjects viewed a brief flash of 8–24 dots either two or three colors randomly arrayed. Their task was to move mouse cursor the centroid (center-of-gravity) each color in pre-designated order. Conventional and idea-detector analyses show that subjects accurately judged all centroids utilizing an astounding 13/24 stimulus dots, with only modest loss accuracy compared judging single-predesignated centroid. The ability concurrently compute is important because it believed judgments are made on salience maps record ignorant features produced salience. Our explanation, instantiated computational model processing, have maps. Dots initially segregated into groups according color, then color-group recorded different map In Part 2, data analyzed terms Attention Operating Characteristics characterize impairments subjects’ color-attention filters (mostly insignificant) encoding efficiency (20% drop for hardest task) making multiple versus single judgments. A new, more sensitive analysis measured five sources subject error variance, four independent, additive variance: imperfect filters; Bayesian-like bias towards central tendency; storage, retrieval, misplacement error; a large residual due mostly inefficient encoding; fifth, interactive source – components increases when judgment required trial. An brain process map, representation relative importance (salience) locations visual space. It needed guide where look next, computing center (technically “centroid”) cluster items, other computations. Here we colors, interleaved, can centroids. As cannot discriminate reporting demonstrates not just one, as commonly believed, but at least