Summary and Conclusions

The purpose of this experiment was to test the hypothesis that a perspective display of aircraft was better than a plan-view display for the task of quickly and accurately identifying descending aircraft. To test this hypothesis, 14 test participants were presented with two separate scenarios, each containing 50 randomly placed aircraft. For each scenario, the participants were asked to identify as quickly as possible all descending aircraft. The number of correctly marked descending aircraft was recorded at five second intervals, and the time to complete the task was also recorded.

An analysis of variance (ANOVA) was performed to determine the statistical significance of any observed performance difference between the perspective and plan-view displays. The ANOVA revealed that there was a significant advantage for the plan-view display when considering the time to fully complete the assigned task. This advantage seems to stem from the extreme difficulty almost all of the participants had in detecting the last 2 or 3 descending aircraft on the perspective display. Another ANOVA was then performed on the same data after the data was truncated to include only the time to detection of 70 percent of the descending aircraft. This analysis revealed that the advantage was now with the perspective display, which, apparently by virtue of direct visual cues, allowed the test participants to more quickly identify most descending aircraft.

From the analysis of the results of the experiment, we conclude that a perspective display is superior to a plan-view display when the operator is required to detect only a majority of the total number of descending aircraft. If the task requires the identification of all descending aircraft, the plan-view display which better facilitates the methodical technique of an exhaustive search is the superior display. Thus, both display types have their strengths and weaknesses with neither clearly superior in general for the task we examined.

The difficulty in finding the last 2 or 3 descenders in the perspective display suggests that an additional cueing feature independent of symbol size may be necessary to draw attention to tactically significant aircraft that may otherwise go unnoticed because of symbol size or clustering. A similar feature would probably benefit the plan-view as well.

To gain further insight into the situations where one display or the other has an advantage for the accomplishment of a particular task, we would recommend further investigation in the following areas.

  1. The number of scenarios should be increased to better understand the effect of the total number of aircraft and the number of aircraft of interest on the system operator's performance.

  2. The effect of the operators' experience level on their performance should be investigated. A few of the test participants stated they would have performed much better with the perspective display if given more time to practice with the system.

  3. The ability to change the operator's point of view in space is a key feature of the FTEWA system. During this experiment, the point of view was fixed because the input device for moving in three-dimensions (the Spaceball) was too novel and would have distracted from the primary objective of the experiment. Further study should be devoted to investigating the usefulness of moving the operator's point of view in three-dimensions in disambiguating aircraft of interest.

  4. In this experiment, the participants had the ability to "hook" aircraft on the perspective display and receive the same type of textual display of aircraft parameters that was available on the plan-view display. One disadvantage of this particular test setup was that the text window was only displayed on the plan-view monitor, even when the "hook" was made from the perspective display. This forced the participants to look back and forth between the two displays instead of allowing them to concentrate on just one display. Future implementations of the FTEWA system may benefit from having the text window available on either display or popping up text information along side hooked tracks.

  5. The scenarios for this experiment were randomly generated and therefore did not necessarily match any realistic scenario that naval commanders would likely encounter. In particular, aircraft did not perform any maneuvers, but were just coasted based on initial flight parameters. This resulted in an implausible, static scenario. Performance should be investigated for the same task when using more realistic scenarios.

Continue

Table of Contents