While performing ongoing tasks pilots are often required to monitor the status of other tasks. Some tasks, such as the requirement to report passing through an altitude, previously discussed, involve monitoring for an event that is known will occur. In other situations, pilots must monitor for events that occur infrequently, if at all. For example, when flying in visual meteorological conditions, pilots must scan outside the cockpit windows for other airplanes that might be on a conflicting path. This may seem a topic of more interest to attention researchers than memory researchers, but in fact pilots report becoming preoccupied with ongoing tasks and forgetting to monitor the status of other tasks for dangerously long periods (Dismukes et al., 1998). The problem is probably greatest when high workload preempts limited resources: Bargh and Chartrand (1999) argued that conscious control of behavior (which monitoring presumably requires at least to some extent) is a very limited resource. However, lapses in monitoring also occur in low and moderate workload situations in which enough time exists to switch attention back and forth between the ongoing task and monitoring and to perform both tasks adequately.
In a flight simulation study, pilots' scanning outside the cockpit increased substantially when potentially conflicting airplanes started appearing, but returned to near baseline levels minutes after the last airplane appeared (Colvin, Dodhia, & Dismukes, 2005). We speculate that it is difficult to maintain the monitoring task goal in working memory when the result of each inspection of the monitored scene reveals that no event has occurred. In this sense the monitoring aspect of the pilots' dual tasks somewhat resembles vigilance tasks (Parasuraman, 1986). Apparently humans are wired to allocate attention heavily toward sources of high information content, and thus have difficulty maintaining monitoring for low-probability events, even when those events may have high consequences (see Wickens, Goh, Helleberg, Horrey, & Talleur, 2003, for a model of attention allocation among tasks). However, this sort of monitoring differs from traditionally studied vigilance tasks in that the pilot must interrupt an ongoing task and shift attention to the thing being monitored. When the pilot goes too long without shifting attention, the monitoring task may slip from working memory, and then must somehow be retrieved, just as in other types of prospective memory situations. This conclusion is supported by a study in which Einstein et al. (2005) found that the level of monitoring declined over the course of the experiment. (The level of monitoring was inferred from the cost to response time of an ongoing task.)
Much of prospective memory laboratory research has focused on stimulus-driven responding, and only a few studies have addressed monitoring as a prospective memory task. Park, Hertzog, Kidder, Morrell, and Mayhorn (1997) required participants to remember to make a response at either 1-minute or 2-minute intervals while performing an ongoing working memory task. Participants were allowed to check elapsed time during the intervals. Monitoring performance was worse with 2-minute intervals, suggesting difficulty maintaining the monitoring task in working memory.
The six-element task developed by Shallice and Burgess (1991) and adapted for prospective memory studies by Kliegel and colleagues (Kliegel, McDaniel, & Einstein, 2000) shares some features with cockpit monitoring tasks. The six-element task requires participants to remember to switch tasks on their own as a function of how far they have progressed on the current task, rather than remembering to switch when a prearranged cue is perceived or at a predetermined interval. Similarly, many cockpit monitoring tasks also require pilots to remember to switch attention periodically from an ongoing task, without the benefit of any specific cue. In both the six-element task and cockpit tasks, the individual's perception of passage of time presumably plays a role, although the mechanisms of this are not understood (see Cicogna, Nigro, Occhionero, & Esposito, 2005, for theoretical speculation about mechanisms of time-based prospective memory).
Many interesting research questions about performance of monitoring tasks of this sort invite study. When the goal of monitoring slips from working memory it eventually reappears—its reappearance triggered by the passage of time, happenstance cues, the context of the cockpit environment, performance of steps of the ongoing task, or something else altogether. What is the relation of the goal and action structures of the ongoing task and the monitoring task? What role does task importance play when monitoring for low-probability, high-consequence events? Do individual differences and situational factors such as fatigue and stress affect monitoring performance? Could pilots and other individuals learn techniques to reduce vulnerability to lapses in monitoring?
Was this article helpful?