Making Sense of Multivariate Data: An Iterative Design of the Visualization and Interactions with Parallel Coordinates

Abstract: There exists today vast accumulations of data that we don't always know what to do with. Qlik is a company that specializes in the visualization of data and specifically in being able to offer as many different types of charts as possible to their customers. Parallel coordinates is a chart that efficiently can visualize multivariate data and this master thesis is about developing such a chart for Qlik, using techniques from interaction design and with a focus on usability. Using techniques like user interviews, focus groups and usability tests the needs of the user were elicited and a parallel coordinates diagram could be developed. The results showed that a parallel coordinates plot could be very effective in visualizing multivariate data but also that it might not always be the most optimal chart for all data sets. The resulting product had basic functionality with some additional tools for further analysis. In this project it was established that there are many factors to think about when deciding how the data should first be presented for the user, which tools that should be implemented and how the user should interact with said tools. Some of those factors may be, who the users are, their background, level of competence and type of data they wish to visualize.