The effectiveness of the space-time cube in the analysis of movement data : the case of Tallinn suburban commuters

The STC is a graphic representation proposed by Torsten Hägerstrand to study spatio-temporal characteristics of movements. It is known that this three-dimensional visual representation can get easily cluttered. However, it is expected it has its role to play in the exploration of movement data when used in combination with other graphic representations. In order to judge its effectiveness in such environment, an extensive usability study consisting of several phases was designed. In this study, a week of movement data of over 203 suburban commuters in Tallinn have been used as test case. The data set has been subject of study by human geographers before. However, a way to explore and capture the daily and weekly rhythms of the commuters was difficult or even lacking.
A first step was a user requirement analysis based on research questions of human geographers. Based on these requirements, a visual environment of multiple coordinated views was developed, and again submitted to the experts to make sure they could apply their own workflow. Besides, specific cartographic designs where applied based on a focus group study with visualization experts. This was the base for a final evaluation of the STC’s performance in a highly interactive geovisual analytics environment, extracting measures on effectiveness, efficiency and user satisfaction.
During two identical experiments with two different groups of users different qualitative and quantitative evaluation methods were applied. The first group consisted of expert users, e.g. human geographers as owners of the problems, and the second group consisted of non-experts, aware of the STC but unfamiliar with the data. The findings of the experiment show significant differences in usability metrics between both groups. The experts were able to apply tools in the developed environment more effectively and efficiently due to their interests to explore familiar places with different movement. Besides, via the different evaluation methods applied, such as eye tracking and think aloud the experiment revealed differences in usability metrics of individual components of the visual environment. To answer specific time related questions the STC was appreciated as the most effective tool.