3D Cities

The experiment contained maps of the cities, where 3D effect was illustrated by 3D buildings. Eye-tracking study was performed to compare user perception of these two types of visualization in maps of the cities. Set of static image stimuli was used for the experiment. Study was performed in within-subject design. For the avoidance of the learning effect, stimuli were modified a bit. Point symbols are placed in different locations, but their number is the same for each pair of stimuli. For evaluation of results, statistical analyses of selected eye-tracking metrics were used, as well as the visual analytics of eye-tracking data. Eye-tracking experiment was enhanced with a questionnaire, examining the respondent’s subjective opinions about each visualization technique.

Methods

Procedure

At the beginning of the experiment, respondents fill out a short questionnaire. Then, the 9-point calibration was performed. Eye-movement recordings with deviation smaller than 1° were included in the experiment. After the calibration, welcome screen with instructions was displayed. The instructions included also the sample question. Respondents’ task was to find out one particular point symbol in the map as fast as possible and mark it with the mouse click. Study was performed in within-subject design. The experiment contained 18 static stimuli with 2D and 3D maps of cities. The stimuli with maps were presented in random order. To unite the starting point of the eye-movement trajectories, the fixation cross was displayed for 500 ms before the stimulus. Respondents had maximum time of 30 second to find the target, but for the most of the tasks, the time was fully sufficient.

Independent variables

The experiment contained screenshots of different internet map portals. These maps were complemented with point symbols. Two types of maps were used – the first one was standard map with buildings represented by polygons, second contained 3D (2,5 D) visualization of buildings.

Stimuli

Maps from three different sources were used. The first one is well-known Google map (stimuli 1-5). Second type of maps contained maps from OpenStreetMap.org (stimuli 6-8). Last stimulus (map 9) also uses OpenStreetMap data through project F4Map. 

 Participants

Total of 40 participants (24 females, 16 males) attended the eye-tracking experiment. Most of them were representatives of academic staff and students. Respondents were originated from different fields. Some of them were cartographers, some of them were not. Majority of participants were 20-25 years old. Participants were not paid for the testing. Before the experiment, respondents filled out the short questionnaire with personal information. Apart from elementary information, like age or sex, they had to answer the question, how often are they using the internet map portals like Google Maps or OpenStreetMap. Most of the participants use the web map portals every day.

Apparatus

For the study, remote eye-tracking device SMI RED 250, developed by SensoMotoric Instruments, was used. This device was operating at frequency of 120 Hz. Eye-tracker was supplemented by web camera, which records participant during the experiment. This video helps to reveal the cause of missing data, respondents’ reactions to the stimuli and their comments to the particular maps. For data visualization and analyses, three different applications were used. First one is SMI BeGaze, which is software developed by the manufacturer of the device. Open source software OGAMA and CommonGIS developed at Fraunhofer Institute in Germany were used for visual analytics of eye-tracking data.

 Analyses

For statistical analysis of eye-movement data, several eye-tracking metrics were calculated. For all metrics, median values for 40 respondents were calculated. Median was used instead of mean because the data had not normal distribution and median also filter out the extreme values. Data was analysed with the use of the Wilcoxon rank sum test and statistically significant difference between 2D and 3D maps was observed on the significance level α = 0.05 in all cases. Analysed metrics were Time to Answer (click), Fixation Count, Fixation Duration Median and Scanpath Length. 

Statistical analysis showed that there are statistically significant differences in eye-tracking metrics between 2D and 3D variant of a particular map, but the results did not indicate that one of the variants is better than the other. Within the analysis of the entire dataset as a sum of all maps, no statistically significant differences were found for any of the studied eye-tracking metric.

For data analyses, two methods introduced by Andrienko et al. were used. First method, Flow Map, represent results of discrete spatial and spatio-temporal aggregation of trajectories. Arrows represent multiple movement of gaze from one location to another.  The thickness of arrows is derived from variable Number of moves between defined voronoi polygons. Only arrows representing more than three moves are displayed. Second used method is Temporal View of Trajectories. The horizontal dimension of the graph represents time, and the colour of the lines displays the distance between current gaze position and the target in pixels.  

Conclusion

No significant differences between 2D and 3D maps were found for four metrics (Time to Answer, Fixation Count, Fixation Duration Median and Scanpath Length). Point symbol search was more difficult on the map where 3D effect was created with use of map tilt (map no. 9). For this map, statistically significant differences were observed for all recorded eye-tracking metrics. This type of maps should not be used very often, because users have problems with orientation in this map.    

On the other hand, results for all other stimuli (maps no 1 – 8) indicate that in situations when it is reasonable and desirable, the 3D map of the city could be used instead of the standard two-dimensional. In the three-dimensional map, more information is contained and the 3D representation did not influence the reading of the map and its comprehensibility.

Outputs

Popelka, S., & Doležalová, J. (2015). Non-photorealistic 3D Visualization in City Maps: An Eye-Tracking Study, Modern Trends in Cartography (pp. 357-367): Springer.

 Downloads