Abstract
It is widely believed that there are two types of spatial attention: voluntary and involuntary attention. We expand this taxonomy by demonstrating that there are at least two mechanisms of involuntary attention. The first is a serial search mechanism and is related to finding the target in a display. In a spatial cueing task, observers have a tendency to search for the target beginning at the cued location (valid trials). If the target is not at that location, observers then search other locations (invalid trials). Hence RTs are faster on valid than invalid trials. The second mechanism occurs at the response decision stage. A spatial cue primes a response to any stimulus that appears at the cued location. This mechanism can be characterized by a formalism called the accumulator model (Usher & McClelland, 2001). The serial model predicts that as the number of display positions increases, the cueing effect will increase. The accumulator model makes the opposite prediction. In most tasks, observers must both find the target (serial search) and decide which target was present (accumulator model). Which of the two mechanisms will limit performance will depend on whether finding the target or discriminating the target is more difficult. We varied the number display positions in a spatial cueing task. We found that when there were no distractors the cueing effect decreased as the number of display positions increased, supporting the accumulator model. When there were distractors, the cueing effect increased as the number of distractors increased, supporting the serial search model. Thus involuntary attention is mediated by at least two different mechanisms: serial search and response decision. We've begun to explore whether different involuntary attention effects (e.g., inhibition of return, contingent capture) are caused by the serial search or decision (accumulator) mechanisms.