Data visualization can be defined as the visual communication of information. As data visualization designers or storytellers, we go through many steps in order to communicate the intended insights in data to an audience. We begin with the data and some goal or intent. We process this data using a variety of methods to discover the relationships contained in the data. Often we find that this analysis by itself, is not enough for us to realize our intentions or our goals. To get closer to our intentions, we have to visualize the data and the underlying relationships. Through the pairing of analysis, interaction, and visualization, we are able to manipulate data into a language imbued with meanings more than what is literally presented. Visualizations impact our perceptions of the data and the underlying insights or narratives. We are able to see causal relationships, uncertainty, and evidence to support or disprove hypotheses.
The complexity of the conversations brought out through data visualizations has steadily increased. Infographics and interactive visualizations are now commonly employed for conveying information to the general public (e.g., as seen in NYT, Reuters, The Pudding, etc.). One important barometer for the success of visualization is whether the intents of the communicator(s) were faithfully conveyed. This intention can be of many forms such as to persuade, to educate, to inform, or even to entertain. A second important factor is that the visualization must be consumable and enjoyable. Finally, it must preserve scientific integrity.
The processes of constructing and displaying visualizations have been widely studied by our community. However, because of the lack of consistency in this literature, there is a growing acknowledgment of a need for frameworks and methodologies for classifying and formalizing the communicative component of visualization. The widespread usage and increasing complexity of data visualizations have increased the importance of delving deeper into this domain of research. In my dissertation research, which consists of four components, I seek to leverage lessons from storytelling and linguistics to offer ideas on how the visualization community and designers can be more precise, consistent, engaging, and impactful in their visual communication of information, particularly when communicating to the public.
An important consideration when designing a visualization tool for non-experts to understand and explore complex data is aesthetics. Visual metaphors provide desirable aesthetics to viewers by employing familiar representations. In the first component, I review several data stories of mine that utilized different visual metaphors. I present different techniques for producing visual metaphors and discuss their strengths and limitations. In the second component of this research, I present a retrospective analysis of a complex visualization that we developed for a science museum. We identified a set of considerations that must be taken into account while designing narrative visualization that seeks to convey complex scientific information in an informal learning environment such as the museum. This work also identified directions for further research into storytelling structures as they apply to data visualization.
The third component is concerned with data characters. In data stories, numerous methods have been identified for constructing and presenting a plot. I posit that there is an opportunity to expand how we think and create the visual elements that present the story. Stories are brought to life by characters; often they are what make a story captivating, enjoyable, memorable, and facilitate following the plot till the end. In an effort to improve how data stories can be produced and be more accessible to larger audiences, I offer guidance on the design of data stories from a character-oriented perspective. Finally, I propose to deepen the understanding of how our intentions as designers affect the visualizations we create by translating linguistic practices and techniques into communicative visualization. I then seek to take these methods and concepts and apply them to a system for automated storytelling. This work enables the integration of theories and frameworks from linguistics and storytelling into visualization and grows our theoretical toolset for studying visualization. Additionally, it illustrates how to apply these introduced frameworks when creating data visualizations with communicative intent. The frameworks and design solutions presented in this dissertation are expected to be applicable to data storytellers or visualization designers who seek to be more effective in communicating their intended messages to an audience.