Designing Effective Visualizations

1. Key (legend) and labelled axes essential to interpretation
2. Use ``intuitive'' mappings where possible (spatial to spatial), though sometimes the non-intuitive mappings can reveal interesting features
3. Use color with care - be aware of context-sensitive color expectations, provide ready access to alternate color maps, user customization
4. Provide easy methods for view selection and modification
5. Avoid overcrowding images - provide users with opportunities to enable or disable features
6. Avoid distorting data (viz lies)
7. Scale your data with care, and convey scaling in key
8. Don't compare apples and oranges - e.g. the correlation between sun spots and the stock market
9. Be concise - avoid excessive gimmicry (e.g. 3-D graphics for 1-D data)
10. Avoid dependence on absolute judgements - relative is more reliable
11. Differentiate original data from derived (e.g. interpolated, smoothed) data
12. Don't forget aesthetics (visualizations should be appealing to the eye)

• Examples of Good and Bad Visualizations