Data Visualization Design
In this workshop we will approach data visualization as a design problem, and learn how to solve this problem systematically. You will learn cognitive and design principles that help you understand what works and why, and how to explore different design solutions more efficiently.
Good data visualization will help you tell the story of your research and make your papers and presentations stand out. Yet, most researchers are never really taught how to visualize data. Instead, their design process is based on intuition, copying others, and try-and-error. This is a waste of time and often produces ineffective results.
In this workshop we will approach data visualization as a design problem, and learn how to solve this problem more systematically. You will see why visualization is a powerful way to communicate your data. You will learn design principles that help you understand what works and why, and practice hands-on how to explore design solutions more efficiently.
- Learn about the value of data visualization and when and why to visualize data.
- Learn about the hidden thinking behind good data visualization: understanding the context of the visualization (e.g., audience, presentation format) and how to use editorial thinking to decide what to show.
- Learn how to approach a data set for visualization, e.g., understand different types of data and the implications for visual encoding.
- Learn about different ways to encode data visually, and understand how data can be encoded more effectively by following basic principles of perception and visual design (e.g., signal detection theory, Gestalt laws)
- Understand the different elements of a chart and how to use them to make more effective visualizations (e.g., annotations, color, composition)
We will focus on explanatory data visualizations (e.g., for posters and presentations) and relatively simple types of data that can be found across most disciplines. We will *not* cover exploratory data analysis, bespoke data visualizations for scientific discovery, domain-specific data types (e.g., text, networks, high-dimensional data), algorithms (e.g., dimensionality reduction, clustering), or advanced visualization techniques (e.g., interaction, animation).
No programming experience is required (and no artistic ability is required, either). However, you should be familiar with fundamental statistical concepts and chart types.
The course is for researchers and graduate students who want to communicate their research more effectively, for example in papers, posters, presentations, or to the general public. Generally, it may be useful to anyone who uses data to inform, support decision making, and motivate change.