“Graphical Perception and Graphical Methods for Analyzing Scientific Data” - or how to create data visualizations that tell a story
- 1. “Graphical Perception and Graphical Methods for Analyzing Scientific Data” Or how to create data visualizations that tell a story Papers We Love Berlin ; June 19th ; Juhis ; @hamatti
- 3. Hi, I’m Juhis! Developer Community & Web Dev @ Futurice Founder of Turku <3 Frontend Used to build data visualizations at Chartio @hamatti in Twitter from Helsinki, !
- 4. 0.000 5.000 9.961 4.373 9.649 6.841 4.063 9.911 0.245 6.545 2.321 0.778 0.039 4.373 3.757 9.843 0.476 7.129 3.159 0.351 9.222 2.321 3.455 9.755 1.355 1.577 9.990 5.314 3.159 9.649 8.645 8.423 1.813 1.147 7.129 0.476 1.813 8.853 9.524 2.871 4.686 9.990 9.755 3.455 9.649 3.159 0.351 6.841 8.423 1.355 8.853 1.813 6.841 0.351 2.321 9.222 5.937 9.911 7.939 9.045 7.409 0.618 9.911 4.063 0.089 5.937 2.871 9.524 9.755 6.545 8.187 1.147 7.679 9.222 4.373 0.039 0.010 4.686 1.577 1.355 0.039 5.627 10.000 5.000 @hamattiPapers We Love Berlin
- 5. @hamattiPapers We Love Berlin
- 6. Read the paper "Graphical Perception and Graphical Methods for Analyzing Scientific Data" by William S. Cleveland and Robert McGill bit.ly/pwlb-dataviz
- 7. Input — Output Quantitative & categorical data encode through position, shape, size, symbols and color Graph decode through human visual system Understanding @hamattiPapers We Love Berlin
- 8. “A graphical method is successful only if the decoding is effective. No matter how clever and how technologically impressive the encoding, it fails if the decoding process fails.” Cleveland & McGill, 1985 @hamattiPapers We Love Berlin
- 9. “A graphical method is successful only if the decoding is effective. No matter how clever and how technologically impressive the encoding, it fails if the decoding process fails.” Cleveland & McGill, 1985 @hamattiPapers We Love Berlin
- 10. Levels of recognition 1. Position along a common scale 2. Position on an identical but non-aligned scales 3. Length 4. Angle or slope 5. Area 6. Volume / Density / Saturation 7. Hue @hamattiPapers We Love Berlin
- 11. 1. Position along a common scale @hamattiPapers We Love Berlin
- 12. 2. Position on an identical but non-aligned scales @hamattiPapers We Love Berlin
- 13. 3. Length @hamattiPapers We Love Berlin
- 14. #1 Never use pie charts @hamattiPapers We Love Berlin
- 15. “Pie charts should never be used” Edward Tufte @hamattiPapers We Love Berlin
- 16. 4. Angle / Slope & 5. Area @hamattiPapers We Love Berlin
- 17. #2 Don’t stack – unless it tells a story @hamattiPapers We Love Berlin
- 18. @hamattiPapers We Love Berlin
- 19. @hamattiPapers We Love Berlin
- 20. #3 Graph what you want to tell @hamattiPapers We Love Berlin
- 23. #4 Show the zero @hamattiPapers We Love Berlin
- 28. #5 Don’t get fancy @hamattiPapers We Love Berlin
- 29. @hamattiPapers We Love Berlin
- 30. https://www.thedataschool.co.uk/ellen-blackburn/a-simple-way-to-make-a-radar-chart/
- 31. #5 Don’t get fancy #1 Never use pie charts #2 Don’t stack – unless it tells a story #3 Graph what you want to tell #4 Show the zero @hamattiPapers We Love Berlin
- 32. Extra reading The Visual Display of Quantitative Data, Edward Tufte Information Dashboard Design, Stephen Few Information Graphics, Sandra Rendgen Cartographies of Time, Daniel Rosenberg Visualize This, Nathan Yau @hamattiPapers We Love Berlin