6 102 6. Gaze-contingent processing improves mobility performance 6.3. Results 6.3.1. Primary outcomes Exp. 1 - Obstacle avoidance Several example mobility trajectories are displayed inFigure6.4. The primary results of the analysis of Experiment 1 are displayed in Figure 6.5. The average number of collisions was significantly higher in the gaze-locked study condition compared to the gaze-contingent condition (p=0.014) and the gaze-ignored condition (p<0.001). Furthermore, the average trial duration was significant higher in the gaze-locked study condition compared to the gaze-contingent study condition (p<0.001) and the gaze-ignored study condition (p<0.001). The regression analysis revealed no significant reduction in the number of collisions across subsequent blocks, but there was a significant reduction of the trial durations (14% per block; p<0.001). The slopes for the different study conditions were not significantly different, indicating no interaction effects. Figure 6.5: Primary results of the mobility task in Experiment 1. A) Average number of collisions per trial (N=23, Paired-samples t-test). B) Average trial duration in seconds (N=23, Wilcoxon signed-rank test). *p<0.0167; **p<0.003; ***p<0.0003. Exp. 2 - Scene recognition The results of the scene recognition task are displayed inFigure6.6(panels a-c). The scene recognition accuracy was not significantly different in the gaze-contingent vision condition compared to the gaze-locked condition, but gaze-contingent vision yielded significantly lower trial durations (51 versus 63 seconds on average, p=0.0012). Furthermore, the subjective rating was significantly higher, by more than 1 point on average (on a scale of 1 to 7; p<0.001). The gaze-ignored control condition was characterized by significantly higher scene classification accuracy compared to the gaze-contingent condition and the gaze-locked condition (p=0.0067 and p=0.012, respectively), as well as higher subjective ratings (p=0.016andp<0.001). The trial durations in the gaze-ignored condition were significantly shorter compared to the gaze-locked condition (p<0.001), but not compared to the gaze-contingent vision condition (p=0.061).
RkJQdWJsaXNoZXIy MTk4NDMw