Distinguishing models of multifocal attention: It's a matter of time

Abstract

In recent years, there has been increasing evidence indicating that humans can selectively process information from multiple locations. While the defining characteristic of multifocal attention is spatial, what distinguishes models that account for this ability is a question of the timing of access to the locations. The temporal properties of multifocal attention were studied to evaluate three models: 1) a serial mechanism that rapidly switches between locations, 2) a parallel process that independently and simultaneously extracts information from multiple locations, and 3) a capacity-limited parallel process. In an adaptation of the method used by Carlson, Hogendoorn & Verstraten (JOV, in press), observers monitored two running analog clocks and reported the “time” on both clocks at the presentation of a cue. Our first experiment found that reporting the “time” from two locations as opposed to one required an additional 100ms of processing time, thus providing evidence against independent, parallel access to multiple locations. In a second experiment, we found that this additional time was not a consequence of switching attention between locations (as would be predicted by a serial model), but instead could be attributed to an increase in processing time at both locations. The results of our experiments therefore indicate that multifocal attention is a capacity-limited parallel process.