Lessons from Visualization Research

A collection of principles drawn from evidence-based research:

Visuals make information more memorable
Data visualization can make information more persuasive
Specific titles, relatable objects and redundancy improve memorability
Distortion through data visualization can be deceptive
Including visual elements associated with science, such as graphs, can enhance a message’s persuasiveness
Narrative visualizations work best when combining author- and reader-driven approaches
Crowdsourcing via Amazon’s Mechanical Turk is a viable way to user-test visualizations
The design of visualization can significantly impact an end-user’s interpretation
The visual language is flexible, but conventions emerge around particular arrangements
A “concrete scale” is a technique that can visually express measures of extreme magnitudes or unfamiliar units
Testing different sequences can be an effective way to improve narrative visualization
The Five Design Sheet methodology tests sketches with clients to arrive at usable design
The perception of visualizations, including its accuracy, can be impacted by social influence