|
Abstract
|
Three experiments assessed four types of aircraft cockpit displays. One display was a circular format whereas the other three were vertical formats. Of the three vertical displays, one had a moving pointer, the remaining two had vertical bars whose length indicated engine status.
The stimuli were prepared on 5 × 5 cm slides. Each slide portrayed either 2, 3, or 4 engines using a particular display-type. On any slide, exactly one of these engines was out of tolerance, i.e., above or below a particular level. Half of the slides were prepared with one engine above tolerance, and half with one engine below. Subjects were instructed to view each slide and report which engine was out of tolerance. Response time and errors were the dependent measures.
Experiment One showed that vertical formats with moving bars lead to the best performance, followed by a vertical format with a moving pointer. Poorest performance was observed for the circular displays.
Experiment Two assessed the four display-types under conditions of good or poor discriminability (clear-cut or borderline.) In the borderline conditions, all engine readouts were within 10% of the total display distance from the tolerance line. In the clear-cut conditions, the readouts were at least 25% of the total display distance from the tolerance line. The pattern of results seen in Experiment One was seen in both the clear-cut and borderline conditions of Experiment Two.
Experiment Three assessed the displays over five days of extended practice. Again, the same pattern of results revealed itself. Over the five days, both response time and errors reflected improved performance. However, on day five, errors were not completely eliminated in the circular display-types and in the vertical display-types with moving pointer. Errors were eliminated for the other displays.
The relative efficacy of each display-type are discussed with regard to number of engines, discriminability of readouts, and practice.
Language: en |