Abstract
[Goal] Covert spatial attention (selection of spatial locations without eye movements) increases contrast sensitivity (CS) in sparse and crowded displays by enhancing the target’s representation and/or suppressing distractors. These benefits occur inflexibly for exogenous attention and flexibly for endogenous attention. To constrain visual attention models, for the first time, we systematically characterized how target enhancement and distractor suppression by exogenous and endogenous attention vary across spatial frequency (SF) and eccentricity within the same observers.
[Methods] Observers performed a 2AFC orientation discrimination task. Tilted (±45°) grating(s) were displayed along the horizontal meridian on both sides of a fixation cross. In the Valid conditions, peripheral precues manipulated exogenous attention and central precues manipulated endogenous attention. In the Neutral conditions, non-informative precues distributed attention across the visual field. Response cues indicated the target. In Experiment 1, a single grating with one of 6 SFs (0.5-11 cpd) was displayed at 4 possible eccentricities (0-12°). Five levels of grating contrast encompassed the participant’s dynamic range, enabling estimates of CS. In Experiment 2, 4 gratings were displayed simultaneously with one of 8 SFs (0.5-11 cpd) and at 2 eccentricities (2°, 6°). On each trial, gratings had the same SF and their contrasts were fixed (based on initial threshold sessions) such that Neutral performance was matched across SF and eccentricity.
[Results and Conclusion] In Experiment 1, Neutral and Valid CS were bandpass across SF and declined with eccentricity. In Experiment 2, Neutral performance (d′) was equated across SF and eccentricity. In both experiments, exogenous attention yielded the largest benefits for SFs above the Neutral preferred SF at each eccentricity, and more so for peripheral locations. In contrast, endogenous attention improved a broad range of lower and higher SFs. Our results highlight how these two types of covert attention distinctly shape basic perceptual dimensions across the visual field.