Predictability matters: on the stimulus-driven account of the multiple-cue effect.

To examine the issue of whether attentional focus can split among noncontiguous locations, Wright and Richard (2003) used a multiple-cue display and found the validity effect for each of the multiple cues. However, they argued against the multiple foci account by claiming that the multiple-cue effect they obtained was stimulus-driven. We doubt whether their cues were indeed exogenous as they claimed since their target could appear in any one of the eight possible locations but the cue validity was set at 50%. In this case, the cue was in fact predictive as to the target location. In the current study, we showed that when the cues were designed to be truly nonpredictive (12.5%), the multiple-cue effect was eliminated (Experiment 2). We replicated the multiple-cue effect when the cue validity was 50%, as in Wright and Richard (Experiment 1), and showed that there was a cue-triggered, attentional-orienting effect but not visual search advantage when the cue remained on the target display (Experiment 3). Our results, therefore, highlight the possibility of involvements of top-down controls in the validity effect found for multiple-cue displays and the importance of taking predictability into consideration in the testing of hypotheses derived from attention theories.