Identifying Bottom-Up and Top-Down Components of Attentional Weight by Experimental Analysis and Computational Modeling.

摘要: The attentional weight of a visual object depends on the contrast features to its local surroundings (feature contrast) and relevance one’s goals relevance). We investigated dependency in partial report experiments with briefly presented stimuli but unspeeded responses. task was letters from mixture (targets) digits (distractors). Color irrelevant task, many stimulus displays contained an item (target or distractor) deviant color (a singleton). results showed concurrent effects feature (color singleton vs. nonsingleton) distractor). A target had higher probability being reported than did nonsingleton target, distractor interfered more strongly targets distractor. Measured by use Bundesen’s (1990) computational theory attention, nearly proportional otherwise similar object, factor proportionality that increased strength singleton. This result is explained generalizing equation attention such becomes product bottom-up top-down relevance) component.