When do dangerous distractors interfere with task performance? The roles of attentional scope and primary-task processing demands.

Dangerous stimuli are often assumed to capture attention automatically and impair ongoing task performance, yet empirical evidence is inconsistent. We examined two potential sources of this inconsistency: whether distractors fall within attentional scope under similarity-load manipulations and whether increased set size imposes sufficient processing demands to weaken distractor suppression. Across three visual-search experiments, peripheral distractor images were drawn from three categories: high-arousal dangerous, medium-arousal dangerous, and low-arousal neutral. Experiment 1 manipulated target-nontarget color similarity in six-item displays. Medium-arousal dangerous distractors slowed responses relative to low-arousal neutral distractors under high, but not low, similarity, whereas high-arousal dangerous distractors produced no reliable interference. Experiment 2 manipulated set size (6 vs. 10 items) while maintaining high target-nontarget color discriminability. Increasing set size produced neither a reliable reaction-time (RT) cost nor a set size × distractor type interaction. Experiment 3 applied the same set-size manipulation in a Landolt-C search requiring item-by-item discrimination. Increasing set size produced substantial RT and error costs. In the 10-item condition, both high- and medium-arousal dangerous distractors slowed responses relative to low-arousal neutral distractors, whereas neither produced reliable RT interference in the 6-item condition. Together, these findings suggest that interference from dangerous distractors is more likely to emerge when the primary task broadens the spatial scope of attention and consumes sufficient cognitive resources to weaken distractor suppression.