ABSTRACT
Abstract Displays consisting of a single target letter, located 2° right or left of fixation, or a target letter and distractor located 2° on opposite sides of fixation were preceded (50, 100 or 150 ms) by bar markers adjacent to possible target letter locations. The location of these markers was either unrelated to target position or identity (Experiment 1), 80% predictive of target location (Experiment 2), or 80% predictive that the target would appear on the opposite side (Experiment 3). Large performance decrements occurred when the markers occurred adjacent to a noise letter in all three experiments. Even in Experiment 3, subjects were largely unsuccessful in overcoming involuntary attentional capture when the bar marker signalled a need to shift attention away from its location. The potency of involuntary capture by onset of marker cues, irrespective of their overall predictive value, merits consideration in theoretical accounts of visual attention and in the design of optimal information displays.
