[Back]

M. Cook:
"An Empirical Evaluation of 2D and Interactive 3D Terrain Visualisations for Cycling Maps";
Supervisor: G. Gartner, F. Ledermann; Department für Geodäsie und Geoinformation, FG Kartographie, 2017; final examination: 05-15-2017.

The last two decades have borne witness to considerable growth in the availability of interactive and 3D cartographies. However, questions about their true usefulness and usability remain, with the results of associated research drawing conflicting conclusions.
This knowledge deficit is especially apparent within the design of cycling maps and cycle route-planners. 3D cartography has enabled new forms of terrain visualisation for these use scenarios, but their usability relative to traditional methods of terrain visualisation is not yet fully understood. In response to this deficit, this thesis aimed to assess the relative usability (measured in terms of efficiency, effectiveness and user-preference) of 2D and 3D elevation
visualisations for cycle route-planners.
In order to fulfil this aim, an empirical user-study was conducted with 36 participants. Participants were asked to solve a range of typical cycle route-planning tasks (related to height detection, slope detection and climb estimation), for a variety of 2D (arrow, colour,
elevation profile) and interactive 3D (3D elevation profile and 3D terrain model) elevation visualisations. Study participants also provided feedback on their visualisation preference for each of the tasks. These usability factors were assessed using a digital survey, which allowed inclusion of interactive 3D visualisations within a controlled experimental setting, whilst also automatically logging user response times, user answers, and preference feedback.
The findings demonstrated significant usability differences between individual 2D and 3D elevation visualisations, although it was not possible to broadly state that one dimensionality was more or less `usable´ than the other. Further, the usability of each visualisation was found to be strongly dependent upon the type of route-planning task.
However, the 2D elevation profile was most efficient in the widest range of scenarios, and those (generally 3D) visualisations which demanded interaction or high levels of cognitive processing were least efficient. The results also showed that those visualisations which placed the lowest cognitive load on the user were most effective; this factor appears to have had a greater influence on efficacy than the dimensionality of the visualisations. For all task types, users preferred 2D visualisations, most especially the 2D elevation profile.