HUMAN REAL-TIME PROCESSING LIMITATIONS

摘要

Human real-time processing resources are critical to the control of processes that occur in real-time and have some unpredictability. Highly predictable processes may be automated. To comprehend, and thus react appropriately to, task demands as they unfold we must process information at a rate at least matching the pace of the task. Investigation of such dynamic processing requires that we measure, amongst other things, absolute loadings (the maximum burden at any given moment across the work session), the rate at which actions/reactions must occur and the effects of multi-tasking during the session. This paper describes a number of task simulations that provide a framework for examining workload effects across continuous work periods of an hour. Processing rate and absolute cognitive load were manipulated and the simulations showed that if absolute load never exceeds capacity then error detection rate during multitasking is very sensitive to the real-time processing demands imposed by the task.