The Effect of a Priori Probability and Complexity on Decision Making in a Supervisory Control Task

Abstract

In the present study we investigated how monitoring and fault management in a ship control task are affected by complexity and a priori probability of disturbances. Participants were required to supervise four independent shipping subsystems and to adjust the subsystems whenever deviations occurred. However, in order to apply the correct action, they first had to diagnose the cause of the deviation by requesting further subsystem information. Complexity and a priori probability were manipulated by varying the number of disturbances occurring simultaneously and the disturbance rates over subsystems. In general, the results indicate that the participants ignored the monitoring function when they were diagnosing a disturbance. Results also show evidence for “cognitive lockup”: Despite the possibility of stabilizing additional system faults and, consequently, increasing their time for diagnosis, participants tended not to interrupt an ongoing fault-finding process. Still, large individual differences were found in both the selected strategy and reasoning abilities.

Using Interference Models to Predict Performance in a Multiple-Task UAV Environment - 2 UAVs

“…In contrast to the findings of Kerstholt et al (1996), or Moray and Rotenberg (1986), operators in the current domain were apparently not postponing action on the second arriving task until the first was completed, a behavior that, if observed, would have reflected strict single channel processing. As described above, there is evidence that pilots may be sampling each display at approximately 3-second intervals (on the average).…”

“…Moray and Rotenberg reasoned that people tend to deal with problems serially, rather than switching between tasks. Kerstholt et al (1996) extended this paradigm by introducing an additional behavioral response that would allow participants to stabilize a faulty subsystem temporarily while they dealt with another fault. If the operator ignored this optimal strategy, in favor of resolving the first problem completely before moving onto the second problem, it would stand to reason that humans prefer strict serial processing over more rapid switching even when it leads to suboptimal behavior.…”

Cockpit automation and mode confusion: The use of auditory inputs for error mitigation

“…Predicting trust and use of automation is very difficult, thereby exacerbating the problem of human-centered design. 17 This has resulted in insufficiently designed displays and feedback mechanisms that actually lower operator mode awareness. 18 The monitoring function removes the operator from the situation.…”

“…Previous studies suggest that people tend to focus on one problem area to the exclusion of other systems, working sequentially through problems. 17 The TNO Human A critical interface that falls under both the Liveware and Environment arms is the relationship between the pilot and ATC. These communications have a direct impact on the cognitive processes employed in the cockpit.…”

Collective Mistrust of Alarms

“…Since the early 1980s researchers have investigated alarm mistrust during complex tasks such as aviation (Bliss, 1997), mining (Mallett, Vaught, & Brnich, 1992), ship handling (Kerstholt, Passenier, Houttuin, & Schuffel (1996) and automobile driving (Nohre, MacKinnon, & Sadalla, 1998). Generally, they have found that participants responded slower, less frequently, and less appropriately to less reliable alarms.…”