Attentional Narrowing: Triggering, Detecting and Overcoming a Threat to Safety.

[摘要] In complex safety-critical domains, such as aviation or medicine, considerable multitasking requirements and attentional demands are imposed on operators who may, during off-nominal events, also experience high levels of anxiety. High task load and anxiety can trigger attentional narrowing – an involuntary reduction in the range of cues that can be utilized by an operator. As evidenced by numerous accidents, attentional narrowing is a highly undesirable and potentially dangerous state as it hampers information gathering, reasoning, and problem solving. However, because the problem is difficult to reproduce in controlled environments, little is known about its triggers, markers and possible countermeasures. Therefore, the goals of this dissertation were to (1) identify reliable triggers of attentional narrowing in controlled laboratory settings, (2) identify real-time markers of attentional narrowing that can also distinguish that phenomenon from focused attention – another state of reduced attentional field that, contrary to attentional narrowing, is deliberate and often desirable, (3) develop and test display designs that help overcome the narrowing of the attentional field. Based on a series of experiments in the context of a visual search task and a multi-tasking environment, novel unsolvable problems were identified as the most reliable trigger of attentional narrowing. Eye tracking was used successfully to detect and trace the phenomenon. Specifically, three eye tracking metrics emerged as promising markers of attentional narrowing: (1) the percentage of fixations, (2) dwell duration and (3) fixation duration in the display area where the novel problem was presented. These metrics were used to develop an algorithm capable of detecting attentional narrowing in real time and distinguishing it from focused attention. A command display (as opposed to status) was shown to support participants in broadening their attentional field and improving their time sharing performance. This dissertation contributes to the knowledge base in attentional narrowing and, more generally, attention management. A novel eye tracking based technique for detecting the attentional state and a promising countermeasure to the problem were developed. Overall, the findings from this research contribute to improved safety and performance in a range of complex high-risk domains.

[发布机构] University of Michigan

[效力级别] Industrial and Operations Engineering

[关键词] Attentional narrowing;Industrial and Operations Engineering;Engineering;Industrial & Operations Engineering