Carl Westin

5-6 Results 105 1.80, p = .072), while neither heading nor combined heading and speed commands differed between conditions. The increased used of speed control may be attributed to the visualization of the entire speed-envelope in the TRI SSD. Finally, the lack of separation losses recorded indicate that safety was maintained with both interfaces. 5-6-3 Transparency perceptions In general, questionnaire results supported the trends found in the simulation data. Selected bar charts for six Likert questions are shown in Figure 5-8. Participants found both interfaces helpful for conflict solving, facilitating an understanding for why a solution was suggested (Figure 5-8(e)), while slightly disagreeing with the statement that better explanations were needed (Figure 5-8(f)). Overall, a general preference for the TRI SSD over the the HB SSD was noted. The TRI SSD was found slightly better for determining which aircraft caused which no-go zone (although not significant) and easier to use for combined solutions (Wilcoxon: z = 2.06, p = .040). However, the TRI SSD was also found significantly more cluttered (Wilcoxon: z = 2.46, p = .014). q 9 HB TRI Number of participants 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 0 2 4 6 8 (a) The representation was help- ful in resolving conflicts Y 9 q 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 0 2 4 6 8 (b) I found this representation too cluttered 9 Y q 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 0 2 4 6 8 (c) The representation made it easier to apply combined so- lutions q 9 HB TRI Number of participants 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 0 2 4 6 8 (d) I understand why the system suggested a certain solution q Y 9 Y 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 0 2 4 6 8 (e) I could determine which air- craft caused which no-go zone Y 9 q 1 2 3 4 5 6 7 Strongly disagree Neutral Strongly agree 0 2 4 6 8 (f) I would have liked more in- formation on which aircraft caused which no-go zone F IGURE 5-8: Bar chart showing Likert questionnaire responses ( N = 9).