Abstract
Purpose. In conjunctive feature search, peripheral cuing improves performance rapidly as the cue lead time (CLT) is increased. After reaching a peak performance at around CLT=50 to100 ms, performance gradually decays to a plateau. This non-monotonic performance curve has been interpreted to support a two-component model for the deployment of attention: An automatic, transient and a voluntary, sustained component. In this study, we investigate the contributions of attention, masking, and priming in shaping the non-monotonic performance curve.
Methods. In a conjunctive search paradigm, stimuli were presented at 12 locations uniformly distributed on a circular array at an eccentricity of 2.5 deg. A cue indicating the possible location of the target appeared at CLTs ranging from 0 to 500 ms. A post-stimulus mask appeared following the offset of search items. Three types of cue and two types of post-stimulus masks were used to control systematically their masking effectiveness. Data were also analyzed in terms of cue-target feature congruency.
Results. Our results show that a specific combination of cue-type and post-stimulus mask is required to produce the non-monotonic performance curve. Furthermore, the transient enhancement of performance at CLTs of ca. 50–100 ms depends strongly on the feature-based congruency between the cue and the search items.
Conclusion. Taken together, our results suggest that the deployment of attention is mainly a monotonic process and that the transient enhancement of performance around CLTs of ca. 50–100 ms reflects a feature-based priming effect by the cue on the search item.
Supported by NSF grant BCS-0114533 and NIH grant R01-MH49892.