Developing a Timeline Visualization for Spatio-temporal Data - Revealing Patterns of Derived Change Rates

Abstract

Timelines are often used as the basis for temporal data visualizations due to their intuitive nature of displaying time in a linear way. Many types of data can be visualized with a timeline. One particular option beyond plotting a fluctuating variable is plotting its rate of change. Looking at the change rate of a dataset instead of the absolute rate can potentially reveal patterns within the data that would otherwise have been less clear. In this research, a visualization method is developed with the goal of revealing patterns, by plotting rates of change and combining a timeline with other graph types. The prototype of the visualization developed here consists of a timeline combined with a line graph, representing the change rate, and a circle graph, representing the absolute rates. The prototype is developed in a 5-stage design framework, which consists of a cyclical process containing defining,prototyping and testing elements. The testing of the prototype is performed in the form of usability testing. Ten testing sessions are performed using questionnaires and the think-aloud method. The tests consist of participants completing a number of assignments. The outcomes of the tests are processed through two coding schemes. The tested prototype is effective in revealing three types of patterns in the data: stability, instability, and the difference between a rising positive change and a rising negative change. The prototype does require certain improvements as determined from the coded transcripts of the test sessions and the results of the questionnaire. A few of the suggested improvement are the following: the option to enlarge graphs for better viewing, a change to one of the interactive tools in order to read specific change rate values, and the addition of multiple legends. Suggested future research include a follow-up study that cycles back through the design thinking stages, implementing the suggested improvements.