Mental models, mental workload, and instrument scanning in flight.

摘要

It was proposed that pilots' instrument scanning behavior operates under two levels of control processes in an information processing hierarchy. At the lower level, the scanning process reflects automatized skilled behavior that is organized in terms of a number of functional components. Each component consists of specialized scanning behavior which is involved in gathering information about a specific parameter of the flight situation (e.g., airspeed, vertical speed). The higher-level cognitive processes provide the means by which automatized scanning behavior is modified according to the specific requirements of the flight procedure (e.g., "reduce airspeed", "begin descent"). These processes reflect the pilots' relative "concern" for the different parameters of the flight situation. The pilots' concerns are based on their understanding (i.e., their mental models) of the requirements of the flight procedure in terms of both the specific types of information required (e.g., airspeed, vertical speed) and the various types of stress (e.g., mental workload, anxiety) associated with the procedure. This thesis attempted to determine the validity of modeling in scanning behavior in terms of higher-level control of the lower-level functional components of scanning behavior.;Six military helicopter pilots participated in the study. Factor analysis was used to describe the various functional components of the pilots' scanning behavior as they flew two different flight procedures. Each procedure was distinguished in terms of specific flight requirements (e.g., reduce airspeed vs. maintain airspeed). Factor analysis was also used to derive, from the pilots' ratings of various items, a number of primary factors which characterized their mental models of the flight procedures. These different factors represented the pilots' understanding of the information requirements and the specific types of stress which characterized the two procedures.;It was hypothesized that (1) the factors which comprised the pilots' mental models of the flight procedures would show a significant statistical relationship with the functional components of their scanning behavior, and (2) for both of the flight procedures, the specific relationship between the factors would reflect the major requirements which characterized the procedure. Both hypotheses were supported by the data.