PERCEIVED STRUCTURE AND THE MAINTENANCE OF ATTENTION

When individuals are presented with a task situation for which they must use complex, sequentially available information, they must construct organizations for this information in order to perform the task effectively. For example, organizing incoming information allows the individual to remember more information, and to predict and prepare for information yet to come. The hypothesis tested in the present study is that such organizations have an impact on an individual\u27s ability to maintain attention to a task and resist distractions in the environment. Specifically, it was hypothesized that (1) people can attain complex, hierarchically organized structures for incoming information; (2) that the boundaries between the units highest in the hierarchy constitute major breaks in the processing and integration of information; and (3) that these major breakpoints are times when people are especially vulnerable to distraction. In the experiment, subjects were trained to perceive particular, defined structures in sequences of stimuli. After training, they performed a task in the context of a video game requiring speeded predictions or classifications of stimulus events. Within the sequences which had been learned, information irrelevant to the task was sometimes displayed. Overall, subjects\u27 response times in the classification/prediction task indicated the psychological reality of the structures for them as they produced responses more slowly when near a high level unit boundary. They were also affected by distraction, slowing performance significantly when distractions were present. However, the major hypothesis was not confirmed: Distraction did not affect performance differentially for higher level units. This null finding is made compelling by the tremendous statistical power of the analysis. The results were interpreted in the light of possible alternative hypotheses; notably, that distraction affects performance in a strictly momentary way, unrelated to sequences of information that the individual processes. In addition, a major limitation to the present test of the hypothesis is discussed, and a revised test of the hypothesis is proposed. The potential relevance of the hypothesis to theories of the development of attention is described