Are the warning icons more attentional?

The rapid growth of attention to visual warnings is a representation of the adaptive behavior of humans. However, the ways warning icons attract attention in the cognition context has yet to be clarified. This research aims to investigate cognitive mechanism of warning icons under various perceptual loads. The results of Experiment A, whose average attentional capture effect of the warning icons (69 ms) was significantly higher than that of the ordinary icons (35 ms), show that compared with ordinary icons, warning icons are prioritized in processing under both high and low perceptual loads. Besides, the attention capturing abilities of non-target warning icons are the same under high and low perceptual loads. To isolate the effects of salient visual features and semantics, warning icons in Experiment B are replaced with transposed icons with saliency but no semantics. The attentional capture effect of warning icons is found to be significantly smaller under high load than under low load, so the effect in Experiment A can be attributed to the semantics of warning icons. In Experiment C the icons of negative and neutral semantics without salient frames are used as interfering stimuli, and the RT to the negative icons (823 ms) was longer than both the RT to the neutral icons (780 ms) and to the no interference icons (743 ms) (P < 0.001), which show that negative icons have stronger attention capturing ability than neutral icons. This research verifies that the semantics of icons is vital, and icons with salient visual features and negative semantics can enhance attentional capture effect.