Optimal allocation of cognitive resources to spatial locations.

Abstract

From experiments showing that subjects differentially attend to parts of the visual field, psychologists have inferred a limitation on human visual information processing capacity. The model presented describes an optimal way to allocate a limited quantity of "cognitive resources," "attention" or "mental effort". For this model, the sense of optimality is maximizing the probability of finding a target. In an experiment to test the model, subjects searched for a single target letter in an otherwise blank field. Two probability distributions were used to select a target location for a trial, but only one distribution was used for a session. For both distributions, the fit of the model was quite good. It is suggested that the model provides a promising way to test the assumption limited visual processing capacity.