1. Visual Presentation of Quantitative Data Cliff Shaffer Virginia Tech Fall 2015

2. What is the issue? Most of us work with and present (in talks or publications) quantitative data. Quantitative data are a part of content in online course materials. We want to be effective in getting our message across

3. What is the issue? Most of us work with and present (in talks or publications) quantitative data. Quantitative data are a part of content in online course materials. We want to be effective in getting our message across Minimal goals: Don’t lie

4. What is the issue? Most of us work with and present (in talks or publications) quantitative data. Quantitative data are a part of content in online course materials. We want to be effective in getting our message across Minimal goals: Don’t lie (at least not by accident)

5. What is the issue? Most of us work with and present (in talks or publications) quantitative data. Quantitative data are a part of content in online course materials. We want to be effective in getting our message across Minimal goals: Don’t lie (at least not by accident) Display graphical stylistic competence Like writing stylistic competence: stick to the point, don’t be verbose, etc.

6. Visual Presentation of Quantitative Information Edward Tufte, John Tukey, etc. have popularized better visualization of quantitative data. Statistical information

7. Quantitative Information Graphical displays of quantitative information should have Power, Truth, Grace Show the data Induce viewer to think about substance Not methodology, graphic design, graphical technique, etc. Avoid distorting what they data have to say Present many numbers in a small space Make large data sets coherent Be closely integrated with statistical and verbal descriptions

8. Power

9. 4 Data Sets (Tufte) What is “interesting” about this information?

10. 4 Data Sets (Tufte) What is “interesting” about this information? N = 11, mean of x: 9.0, mean of y: 7.5 Regression equation: y = 3 + 0.5x; Sum of squares: 110.0 Correlation coefficient:.82; r2 =.67

11. Visualize It

12. Data need to have meaning to have power

13. Map-based presentations

14. No Power We can calculate Data Density: # of entries/area of graphic

15. Newspaper Graphics

16. Another Powerful Example

17. Train Schedule

19. Napoleon’s Invasion of Russia

20. Small Multiples

21. Truth

22. Hiding the Truth (1)

23. Hiding the Truth (2)

24. Lying by Accident (?)

25. Misleading (for whatever reason)

26. Facebook Graphic Part 1

27. Facebook Graphic Part 2

28. Fuel Economy Standards (1)

29. Fuel Economy Standards (2)

30. Fuel Economy Standards (3)

31. More Binning Problems

32. Stock Question My stock is worth $1 If my stock goes up by 50%, then down by 50%, am I richer or poorer? Or the same?

33. Ratio Games (1) TestTotalPass% Pass 13006020% 25024% Total3506220.6% TestTotalPass% Pass 132825% 2500408% Total532489% System A System B

34. Ratio Games (2) Purest form AB 1.51 21 Sum1.5 Avt.75 AB 10.51.0 22.01.0 Sum2.52.0 Avt1.25 AB 11.02.0 21.00.5 Sum2.02.5 Avt1.01.25

35. Class Size (1) Claim: “90% of our classes are small. Scenario: 1000 students Each must take 10 classes to graduate 10 classes with infinite enrollment capacity (up to 1000) 90 Independent Study classes w/ 1 student each

36. Class Size (1) Claim: “90% of our classes are small. Scenario: 1000 students Each must take 10 classes to graduate 10 classes with infinite enrollment capacity (up to 1000) 90 Independent Study classes w/ 1 student each Consequence: Of 10,000 “student-class experiences”, 90 are in IS, while 9,910 (> 99%) are in large classes. 90% of the classes given are small.

37. Class Size (2) Claim: “80% of our classes are small. Scenario: Each student must take 10 classes to graduate “Large classes” have capacity 100, “Small classes” have capacity 20. Ratio is 4:1, or 180 seats for a “unit” of 5 classes. Given: 100 classes: 20 Large, 80 Small. Total is 3600 seats, can graduate 360 students.

38. Class Size (2) Claim: “80% of our classes are small. Scenario: Each student must take 10 classes to graduate “Large classes” have capacity 100, “Small classes” have capacity 20. Ratio is 4:1, or 180 seats for a “unit” of 5 classes. Given: 100 classes: 20 Large, 80 Small. Total is 3600 seats, can graduate 360 students. Consequence: 2000 of 3600 seats are in Large classes (55.5%) The average student takes 5-6 Large classes, 4-5 Small classes 80% of the classes given are small.

39. Grace

40. Avoid chart junk

41. Minimize data ink

42. Use of Color