New study: dynamic color variables reduce errors for color-blind users

Design · 4 min read

Researchers tested two approaches across multiple design systems: static contrast-preserving palettes versus dynamic color variables where semantic tokens compute accessible contrast at runtime. Participants with protanopia, deuteranopia, and protanomaly completed UI tasks like form completion and data interpretation. Systems using dynamic variables—where foreground/background values adjust based on luminance relationships—consistently outperformed static palettes, particularly in dense data visualizations and dashboards.

The paper highlights that dynamic tokens simplify maintenance for large teams. Because tokens encapsulate contrast logic, component builders don't need to anticipate every color combination; the system computes accessible swaps before rendering. The study also found that combining dynamic tokens with non-color cues (icons, patterns, text labels) produced the largest improvements in accuracy and confidence among participants.

Practitioners interviewed in the paper urged design systems to provide both runtime token APIs and a clear migration path from legacy palettes. The researchers have published a reference implementation and a checklist that design teams can adopt to measure contrast health across components and make dynamic token adoption incremental rather than disruptive.