Challenges and Pitfalls of Data Visualization

"A picture says more than a thousand words"". This sounds like a true statement. While it's true that it is easier to tell a story with visuals than with words, it makes a huge difference which visual you are using for which story. What is the real story behind a visual. Using the wrong visual to tell a certain story can reduce the benefit of a visual to zero. In this presentation we will go through the challenges & pitfalls of Data Visualization.

Transcript

1. nlOUG Tech Experience 2018 Challenges and Pitfalls of Data Visualization 7 June 2018 - 14:00 - 14:45 / De Rijtuigenloods, Amersfoort
2. Welcome Daan Bakboord http://www.daanbakboord.com https://twitter.com/daanbakboord https://nl.linkedin.com/in/daanbakboord http://blog.daanalytics.nl #obihackers
3. Agenda • Introduction • Mind • Perception • Psychology • Thinking • Dashboard Design Challenges and Pitfalls of Data Visualization
4. Initiatives IT IT IT Reliable Information / Insights A B Change / Policy / Strategy Data
6. A picture says more than a thousand words
8. Visual Perception
9. Visual Perception
10. Visual Perception
11. Visual Perception - Process Register Observe & Interpret
12. Visual Perception • Iconic Memory • letters, colors, digits, shapes, etc. • e.g. Computer buffer • Short Term memory • chunks of information FiFo • e.g. Random Access Memorie (RAM) • Long Term memory • recognise people, events, patterns, etc. • e.g. Computer harddisk
13. “Gestalt” Psychology Overall picture before perception of components Figure / Ground Law of Grouping Whole is more than Sum of Parts
14. It’s all about Perception
15. Gestalt Principles of Perception Closure Proximity Similarity Continuation Prägnanz (Good Figure, Simplicity)
16. Optical Illusion
17. Optical Illusion
18. Two System Way of Thinking • System I • Thinking Fast • Intuitive, “Gut” reaction • Jump to Conclusion • “Always On” • System II • Thinking Slow • Reflection • Problem solving • Analysing
19. Dashboard Design • Present information Logical and Ordered • Signaling should be Clear and Unambiguous • Choose the right Visualization • Don’t give System I the opportunity to go wrong • Optimal balance between System I & System 2 • Focus on the core message
20. Subconscious Attributes • Color (Intensity) • Shape • Positioning • Motion
21. Using Color 98675321098739873947612902 16438902451209137828462010 08345601395660087414383020 47294901837900298265150137 98675321098739873947612902 16438902451209137828462010 08345601395660087414383020 47294901837900298265150137 Count the number ‘5’
23. Color Blind Experience Color Blindness via Color Oracle
24. Right Wrong Using Color Associations B & W Printing Too much color –where to focus?
25. Data Ink Ratio Edward Tufte
26. Data Ink Ratio • Minimize the Data Ink Ratio as much as possible • Increase the Data Ink Ratio if necessary to tell the story
27. Graphs
28. Graphs
29. After a demo by an Account Manager Visuals
30. Report
31. Stephen Few http://www.perceptualedge.com/
32. Stephen Few Bullet Graph SparkLine
33. JQuery
34. See the Signals • Our Brains are better suited to process Images • Turn complex data sets into Easy to Digest and Supporting Insights • Identify interesting Patterns and Opportunities and Previously Unknown Trends
35. Questions?
36. Thanks!