Example-Aided Design: Supporting Visualization Ideation and Implementation

Abstract

Data visualizations are a powerful tool for understanding data and conveying complex information to audiences with diverse technical backgrounds. However, designers often struggle to choose and craft visual representations that fit their analytical and communicative goals. In practice, when designers cannot develop design ideas, they often use examples of past visualizations to find inspiration or conceptualize the space of possible designs, offering an avenue for designers to understand what designs are possible for the data they are working with and how they can author these designs. Yet, many existing visualization authoring tools assume designers always have a fully specified design intent, completely overlooking the use of examples as cognitive scaffolds for the visualization design process. This dissertation systematically investigates example-aided design in data visualization. First, through semi-structured interviews with novice and expert visualization designers, this dissertation studies what, how, and why designers engage in example-aided design. The results of this interview study reveal designers’ motivations and processes for engaging in example-aided design, highlighting gaps between designers’ practices and current visualization tools. Specifically, there is a lack of adequate support for retrieving diverse design examples and enabling designers’ example adoption strategies (i.e., select & merge, replicate & modify, trial & error). Second, this dissertation presents a two-part controlled experiment to identify and measure the factors that modulate curated examples’ influence on design outcomes and investigate the relationship between example browsing patterns and design outcomes. The findings from this experiment show that design outcomes are influenced not only by data schema similarities between examples and the working dataset but also by the timing of example exposure during ideation and the visual complexity of the examples. Furthermore, connections exist between designers’ strategies to filter and browse example galleries and the quantity and variety of designs they produce. Finally, towards translating these findings of how designers engage in example-aided design to improve design tools, this dissertation presents Mirny, a web-based tool that supports example-based iterative prototyping of interactive D3 visualizations, enabling the process of replicating and modifying design examples to suit a designer’s design and analytical contexts. By integrating empirical insights with a practical authoring environment, this work advances our understanding of how to harness examples for more expressive visualization creation and lays a foundation for a new avenue of research on example-aided design for next-generation data visualization design tools.