Wednesday, February 16, 2011

Presenting Quantitative Data Part 2

Last night sorting some pdf files I came across an article from A. Globus and E. Raible call 14 Ways to Say Nothing with Scientific Visualization. How appropriate! Although not all the 14 ways directly apply to presenting quantitative data, some of them do. Here is a quick summary:


1. Never include a color legend. Or for that matter no legend at all. Forgot the labels on the x, y and z axes? Does your data have units? Which ones are they?  Is your title clear and accurate? 

2.  Avoid annotation. Draw line and arrow to point at important features you data shows. This will help the audience understand your data better. But don't over do! Here is an example of a good annotation. During his 2011 State of Union Address Barack Obama used an enhanced presentation on the White House webcast. After 10 minutes and 30 seconds the first graph containing data appears. It is about the recovery of the Dow Jones after he took office. Now, I can say there has been a recovery because of the annotation on the chart. That is the point of annotations.



7.  Never learn anything about the data discipline. As I said in part 1, the books from Tufte and Few are good ways to get started.


8. Never compare your results with others. I also talked about this yesterday and here is another example.  In their State of the Mac October 2010 event Apple's Chief  Operating Officer Tim Cook discusses the financial issues of the Mac putting them in context. Click here to see. Sorry,  Quicktime require.

10. Never cite reference for the Data. No further comments require...right?

And let me add one more that sadly I have seen often

15. Never check how the visualization looks when projected on the canvas. The problem with no checking your charts and graphs is that you don't know if the width and color of the lines will be seen when using a LCD projector. If you had practice and got feedback this shouldn't be an issue.

No comments:

Post a Comment