Quantitative models of attention and response processes in shape identification tasks (Q1329122)

The goal of this paper is to describe two quantitative models of attention that are formulated from current knowledge of brain architecture and circuitry, and then link the attention models to response time models having properties that are similar to the recruiting of neural impulses to fire a cell in a neural circuit. It is hoped that the resulting compound models can provide useful ways to test properties of attention and response generation against experimental data. The plan of the paper is (1) to describe two quantitative models of the attention processes involved when subjects identify a target object in a cluttered field; (2) to describe how the attention and identification processes can be joined with response-counter mechanisms; (3) to describe three response counter models; and (4) to briefly compare predictions made by the attention and response time models with some relevant data. The mathematical concepts underlying both the attention models and the response-counter models are rooted in the stimulus sampling theory of \textit{W. K. Estes} [Psychol. Rev. 57, 94-107 (1950)]. In the present formulations, a stimulus display is conceptualized in terms of the stimulus features residing at specific locations. The stimulus features, in turn, are formally represented as stimulus elements, and the selective attentional processing of a compact set of elements is represented by probabilistic sampling of locations in visual space.