[OT] examples of good and bad graphic practice

[baptiste auguie]

Hi,

I'm compiling examples for a talk on producing scientific graphs in physics and chemistry. I will present 'the good, the bad, and the ugly' as a means of introduction to best practices and mistakes to avoid when making a plot, regardless of the tool being used. Maybe I'll have a few slides with the original bad graph, and a proposed improvement next to it. Of course being a ggplot2 user, I only have vague memories of truly bad and ugly graphics. Would you have a link to share presenting such examples, along the lines of Tufte's books but online?

Things that come to mind in no particular order,

- 3D pie chart

- some 2D pie charts

- 3D barplots

- double-axis

- misleading axes, and scales

- inappropriate colours

- ugly fonts

- useless "decorations" such as shadows, background image, etc.

- uninformative legend

- plot with too much information (wrong type of plot)

- export in lossy jpeg format

- manual tweaking of data points / scales

Cheers, and a Happy New Year!

baptiste

[Thomas]

Some of the Martin Krzywinski talks you can find on http://circos.ca/tutorials/course/ (I think it is session 1) includes various examples of bad graphics taken from the literature - They mostly fall in the 'unnecessary decoration' and 'low data-to-ink ratio'

Happy New Year

Thomas

[Phil Taylor]

Check out all of the material at: http://www.datavis.ca

Phil Taylor.

> --
> You received this message because you are subscribed to the ggplot2
> mailing list.
> Please provide a reproducible example:
> https://github.com/hadley/devtools/wiki/Reproducibility
>
> To post: email ggp...@googlegroups.com
> To unsubscribe: email ggplot2+u...@googlegroups.com
> More options: http://groups.google.com/group/ggplot2

--
Phil Taylor
Bird Studies Canada Chair in Ornithology
Department of Biology, Acadia University
Wolfville, NS, Canada
902-585-1287

(NOTE: All of my various email addresses end up in the same place)

[Brian Diggs]

Here are some links I've collected over time for this purpose:

http://thinkprogress.org/media/2009/11/23/70862/fox-pie-chart/

http://junkcharts.typepad.com/junk_charts/2010/11/rightsizing-the-graph-edition.html

http://junkcharts.typepad.com/junk_charts/2011/07/largest-ever-difference-between-328-and-327-spotted-in-new-york-city.html

http://junkcharts.typepad.com/junk_charts/2009/10/getting-the-nobel-wrong.html

http://www.venganza.org/images/spreadword/pchart1.jpg

And one that tries to be as wrong as possible for the sake of being
wrong (and being funny):

http://cheezburger.com/6214307840

And a few more that use humor to poke fun at bad graphs:

http://www.marriedtothesea.com/index.php?date=042810

http://xkcd.com/833/

> Cheers, and a Happy New Year!
>
> baptiste
>

--
Brian S. Diggs, PhD
Senior Research Associate, Department of Surgery
Oregon Health & Science University

[baptiste auguie]

Hi,

Just a word of thank you to all who responded; I got very helpful links and examples. I gave the talk today and got positive feedback. 

If anyone is interested in seeing the slides, write me off-list and I'll send you a link when they're on the uni's webpage.

Cheers,

baptiste