Abstract
Neurologically normal subjects exhibit a systematic leftward error in estimating the midpoint (PSE) of horizontal lines (Jewell & McCourt, 2000). This tonic leftward error (pseudoneglect) reveals a contralateral bias in the deployment of visuospatial attention by the dominant (right) hemisphere, and is observed in both egocentric and allocentric (object-referenced) coordinate systems. Transient exogenous visual cues automatically recruit spatial attention to cued locations and phasically modulate tonic line bisection error (McCourt, Garlinghouse & Reuter-Lorenz, 2005). However, the optimal cue-line onset asynchrony (SOA), the cue-contrast response of this phasic cueing effect, and the relative effectiveness of cue locations relative to line endpoints are unknown. We report three experiments where subjects performed a tachistoscopic visual line bisection task (McCourt & Olafson, 1997). Pretransected lines (27 × 3°, 150 ms) were preceded by cues delivered to the left and right line ends at a variety of contrasts, positions and stimulus onset asynchronies (SOAs). Cues were circular cosine gratings (3° diameter, 3 c/d, variable contrast, 30 ms duration). Experiment 1 manipulated cue-line SOA (40–500 ms). Optimal SOA was 90 ms, and there was no evidence for inhibition of return. Experiment 2 manipulated cue contrast (0.4–100%). Cue contrasts below 12% were ineffective in modulating PSE, and the maximal effect occurred for cues of 100% contrast. Experiment 3 manipulated the azimuthal location of the cue relative to the line endpoints (8 locations, varying from completely beyond the line endpoints to near line center). Cues at all spatial locations except the one lying completely beyond the line endpoints resulted in significant modulation of PSE. These results suggest that the mechanisms subserving exogenous cueing effects in line bisection have low contrast gain, and that cues exert their effects in an object-referenced coordinate system.