On the information processing dynamics of inhibition of return.

The detection of a target occurs expeditiously when it is closely preceded by an uninformative cue that had occupied the same location as the target. However, as the time interval between the cue and target increases, the detection of targets at cued locations is delayed. This latter effect is known as inhibition of return (IOR), referring to a mechanism that is believed to promote search to novel locations by inhibiting attention from returning to a recently visited location (Posner & Cohen, 1984). Although the name of the effect implies a putative mechanism, the inhibition of attention, there are three other prominent explanations for the IOR effect. The slowed time to detect targets at the cued location may be the result of a raised criterion to respond to the target's location (criterion-shift account), the inhibition of responding to the target (inhibited response account), or a temporary disconnection between stimulus and response stages of processing (stimulus-response disconnection account). The goal of the present investigation was to assess these accounts of the IOR effect. Experiments 1--3 examined the effects of IOR on the speed-accuracy tradeoff function. Whether IOR improved or reduced the sensitivity to the target depended on the context of the task. However, the results consistently showed that responding was more conservative for targets appearing at the cued location when response speed was stressed. In Experiments 4 and 5, the IOR effect was assessed in conjunction with non-spatial expectancies. The resulting interaction between the IOR effect with a non-spatial expectancy suggests that the IOR effect is, in part, a criterion-shift. The results are discussed within an integrated framework that posits that IOR has multiple effects on information processing. Moreover, there are particular task contexts that have a strong influence on the manifestation of the IOR effect.