Abstract: Some authors have recently argued that IOR occurs only in detection and localization tasks, but not in tasks that require target discrimination. Other researchers have consistently shown the effect in discrimination tasks. We argue that these discrepancies are due to the use of different methods for measuring IOR. Those who have failed to show IOR in discrimination tasks have used a target-target method, while those who have shown the effect have used a cue-target method, which seems to be more appropriate to study IOR to a location. However, there are still other task differences which can not be accounted for by differences in method. According to Lupiáñez, et al. (in press), the most important difference regarding IOR in detection and discrimination tasks is the time-course of the effect: In discrimination tasks, IOR appears at longer cue-target onset asynchronies. In the first three experiments of this paper we replicate and extended this finding to a new procedure. In two following experiments we find a similar time-course dissociation between two discrimination tasks with different attentional demands. We argue that the time-course of the IOR effect depends on how long attention is endogenously maintained in the cued location. The more demanding the task the longer attention is maintained, and therefore the later IOR appears. In the last two experiments a distractor was displayed together with the target in order to discourage maintenance of attention, and IOR appeared sooner. The results are discussed in terms of an interaction between the exogenous and endogenous attentional system.

(1997). Submitted: JEP:HPP