This Doctoral Dissertation Improvement Grant supports research by the co-PI to investigate the relationship between task-unrelated thinking, that is, mind wandering and the task that is being executed at the time. The research, focusing on behavior in a cockpit flight deck environment, will examine whether certain task features are related to higher incidence of mind wandering and whether incidents of mind wandering are related to human error. In terms of intellectual merit, the research will develop a theory of the consequences of mind wandering on task performance based on established findings in the existing literature, survey research and cognitive task analysis. The theory will be tested experimentally in the context of a simulated flight control system.
The broader impact of this research is that it will advance knowledge and understanding in Human Factors by defining a theoretical construct for mind wandering and developing metrics for measuring this construct in operator controlled systems. In addition, this research will yield a task-taxonomy to explain when mind wandering is most likely to occur and suggest new methods to periodically re-engage operators in ways that minimize fatigue, increase situation awareness, and improve performance in operator controlled systems.
Humans often operate complex systems to complete critical tasks safely. Cognitive factors such as attention, emotion, and mental fatigue can negatively affect the operator’s ability to stay focused on the task and make timely decisions. This research project employed both qualitative and quantitative methods to study another cognitive factor -- mind wandering -- and its effects on the performance of drivers and pilots. Study I concerned the definition of mind wandering. It used a factor scenario survey to investigate how pilots define mind wandering as a function of the timing of an off-task thought (thinking about future vs. past events), the distance in time between the current task and the events in the off-task thought, and how related the off-task thoughts are to the current task. The results from the 39 pilots who completed the survey indicated that, in general, the less closely related the off task thoughts are to the current task, the more likely they were to consider that off-task thought to be mind wandering. In addition, the timing and relatedness of off-task thoughts had joint effects. Off-task thoughts that were unrelated to the operator’s task was always likely to be considered to be mind wandering whether it was a future event or a past event or whether it was close or distant in time. However, if an off-task thought was related to the task, pilots might still consider it to be mind-wandering if the off-task thought was about either a past event or an event that occurred close in time to the task. Study II examined the relationship between performance decrements and types of off task thinking, including thoughts that were relevant to driving but not to the specific aspect of driving currently being performance (slightly off task-thoughts), thoughts that were unrelated to driving (mind wanderings), and thoughts unrelated to driving and being unaware that the mind wandering was occurring (wanderings without meta-awareness). Drivers were given brief question probes five times during the driving task in which they rated their thoughts as on-task, slightly off-task, mind wandering, or mind wandering without meta-awareness. Driving performance was measured by amount of time exceeding the speed limit and number of deviations out of the driver’s lane during the 14 seconds prior to the probe. One hundred eighteen undergraduate students with driving experience participated in the study. The highest mean number of lane deviations and the second highest mean seconds of speeding occurred when drivers were mind wandering without meta-awareness. In contrast, when drivers were engaging in on-task thinking, they did the least speeding and the fewest lane deviations. Also, drivers were most likely to mind wander during parts of the driving task in which the cognitive workload was relatively low. The proportion of off task thoughts that were classified as mind wandering without meat-awareness did not differ much as a function of cognitive workload. In Study III, 17 general aviation pilots performing a simulated flight task described their on-task and off-task thoughts in response to periodic question probes and on a post-flight simulation survey. Fifty-two percent of their reported thoughts were on task, related either to flying or to navigation. Twenty-nine percent of their reported thoughts were relevant to the current task, but were temporally off-task in that they involved thinking about past or future events, including planning for upcoming steps (17%) and revisiting past flight performance during the simulation (12%). Nineteen percent of the thoughts were unrelated to flying, with 8% involving mind-wandering triggered by external events and 11% involving mind wandering with an apparent internal trigger. The frequency of these uncued, spontaneous thoughts was related to the level of stress in a pilot’s life, as measured by the Stressful Life Event inventory. These studies indicate that, during tasks like driving or flying, people’s thoughts are often off the present task. Some of the off-task thinking may still be relevant to the task but focused on related past events or planning for future events. However, a substantial amount of thinking is unrelated to the task and may be evoked either externally or internally. Mind-wandering was less likely when the participant’s minds were highly occupied due to the complexity of the task. And, when people mind-wandered, they were more likely to make errors. Mind-wandering could be considered to be like multitasking or task interruption except that the secondary task of mind wandering is covert and often self-initiated.
