Static and Dynamic Visual Mappings to Explore Bivariate Data Across Time

Abstract

Time and time-oriented data are pivotal for many Visual Analytics scenarios. In principle, the time dimension can be represented using the display space (static mapping, e.g., line plot) or physical time (dynamic mapping, e.g., animation). Dynamic mapping is more direct and, thus, a promising visual metaphor. However, there are contradicting results on its effectiveness for analytical reasoning. To explore the mutual behavior of two variables, the most popular visual method is probably the scatter plot. However, an individual scatter plot is only a static snapshot of the relationship of two variables and developments over time cannot be seen. Hence, time can be added either using an animated scatter plot, or by repetition (small multiples). Moreover, with small multiples either states or changes over time might be emphasized. To gather empirical evidence about the advantages and disadvantages of these different approaches, we developed a Visual Analytics prototype and conducted a comparative user study.