Data Visualisation Design (LATCH)

Description

A stepwise practice for building a visualisation: choose organisation (LATCH), coordinate system, scales and one visual cue per data type, keeping form fitted to content.

The practice treats a chart as a sequence of explicit design decisions. First decide which relationships matter. Then choose an organising principle from Wurman's LATCH (location, alphabet, time, category, hierarchy), a coordinate system (Cartesian, polar, geographic) and scales (linear, categorical, percentage, logarithmic, ordinal, time). Finally assign one visual cue per variable from the set position, length, angle, direction, shape, area, volume, saturation and hue, preferring the cues Cleveland and McGill found people judge most accurately (position and length) for the most important comparisons. Layering one cue per variable lets relationships show without clutter. Decorative choices can mislead, and area and volume are hard to judge, so the result should be tested by having others read it.

Attributes

Loss type
Active
Effort
Hours

Steps

  1. Decide which relationships in the data the viewer must see.
  2. Pick an organising principle (location, alphabet, time, category or hierarchy) and a coordinate system.
  3. Set scales appropriate to each variable.
  4. Assign one visual cue per variable, using position or length for the key comparison and avoiding area and volume for precise values.
  5. Remove decoration that does not encode data.
  6. Test readability by asking two or three people to say what they see first and what the main message is, then revise.

Sources

  • Universal Methods of Design (Martin & Hanington, 2019)
  • Wurman, R. S. (1989). Information Anxiety. New York: Doubleday.
  • Cleveland, W. S. and McGill, R. (1984). Graphical perception: Theory, experimentation, and application to the development of graphical methods. Journal of the American Statistical Association, 79(387), 531-554.
  • A-Z of Digital Research Methods (Dawson, 2020)
  • The SAGE Handbook of Visual Research Methods (Margolis & Pauwels, 2011)
  • The Principles of Writing in Psychology (Smyth, 2004)
  • Everyday Information Architecture (Martin, 2019)
  • Designing Interfaces (Tidwell, 2010)
  • A Practical Guide to Designing with Data (Suda, 2010)

Related talks

Also known as

  • Information visualisation for research findings
  • LATCH organisation
  • Visual encoding choice

Categories

  • Design
  • Design & Product Practice