Recent research with visual search tasks has suggested that stimuli which appear as sudden onsets (new objects) have attentional priority over stimuli that are created by removing segments of premasks (non-onset stimuli). Attentional capture by sudden onsets occurs despite the fact that the appearance of these new objects predicts neither the identity nor the location ofthe target in the visual search task.In three experiments, we examined the extent to which attentional capture by sudden onsets could be modulated by offset transients used to create non-onset objects. To that end, we systematically manipulated the ratio of non-onset to onset stimuli in the display (display ratio) as well as the ratio of offset to onset segments between the stimulus types (stimulus ratio). Increases in either the stimulus ratio or the display ratio resulted in increases in the visual search slopes for the onset targets. These results suggest that the ability of sudden onsets (new objects) to capture attention is influenced by stimulusdriven factors, such as environmental change. Interestingly, the results also indicated that goal-directed or purposeful search for sudden-onset (new-object) targets was relatively uninfluenced by the amount of change in the visual display. Therefore, it would appear that environmental change has differential effects on goal-directed and stimulus-driven search. These results are discussed in terms of their implications for our understanding of attentional capture.

In two experiments we examined a number of related factors postulated to influence head-up display (HUD) performance. We addressed the benefit of reduced scanning and the cost of increasing the number of elements in the visual field by comparing a superimposed HUD with an identical display in a head-down position in varying visibility conditions. We explored the extent to which the characteristics of HUD symbology support a division of attention by contrasting conformal symbology (which links elements of the display image to elements of the far domain) with traditional instrument landing system (ILS) symbology. Together the two experiments provide strong evidence that minimizing scanning between flight instruments and the far domain contributes substantially to the observed HUD performance advantage. Experiment 1 provides little evidence for a performance cost attributable to visual clutter. In Experiment 2 the pattern of differences in lateral tracking error between conformal and traditional ILS symbology supports the hypothesis that, to the extent that the symbology forms an object with the far domain, attention may be divided between the superimposed image and its counterpart in the far domain.

We examined the ability of younger and older adults to selectively process moving items and ignore stationary items in a task that required the search for a target defined by a conjunction of movement and form (i.e., search for a moving X among moving Os and stationary Xs) in displays of 5, 9, 17, and 25 stimuli (Experiment 1) and displays of 5, 10, and 20 stimuli (Experiment 2). We also investigated subjects' performance in two feature search tasks, the search for a target defined by movement or form. Finally, we examined the influence of practice on feature and conjunction search. Younger and older adults searched the displays at similar rates in the feature and conjunction search tasks. Older and younger adults also benefited equivalently from practice. These data suggest age-equivalence in the processes which underlie feature search in dynamic environments as well as those processes responsible for the segregation of moving and stationary objects in the visual field.

The present study evaluated dual-task performance as a function of the vertical separation between a tracking task and a discrete-response task, to provide data relevant to the positioning of aircraft head-up display (HUD) information. The data were consistent with Sanders” (1970) research on visual scanning where a nonlinear decrease in performance as a function of the horizontal separation between two displays was observed. Performance is equivalent across a range of visual angles from superimposition to 6.4° vertical separation between displays. The cost to performance is increased for moderate vertical separations (9.6° to 22.5°) where visual scanning is required. At larger separations, the performance cost increases linearly with visual angle, where head movements may begin to supplement eye movements in order to access information. The function which describes the cost of vertical separation was observed to be larger at both small and moderate visual angles when the information in the two displays required integration. The data suggest that nonconformal HUD information may be placed a few degrees down from a superimposed position without a significant performance loss.

The appearance of a new object within a multiple item display has been shown to capture attention in a stimulus-driven manner. Capture may be either beneficial or detrimental to performance depending on whether the new object is a target or distractor. In the present study we show that the ability of new objects to capture attention is mediated by the number of objects that change or morph. This finding establishes a boundary condition of the phenomena of attentional capture and has implications for the design of complex displays.