up vote 6 down vote favorite
1

The photo stack exchange site has a weekly submission (by community members) for a photo of the week. I believe we could do a similar process, where community members submit a data visualization image to the site, the community votes on it, and the winning submission is posted on the newly created blog.

It doesn't have to be weekly, nor does it have to conform to any of the conventions that the photo site has made for their photo of the week (although many offhand seem reasonable to apply here, and I think would be a reasonable comparison to see how it would work here on this site). I'm simply asking to see if the community is interested in partaking in such a project, and if so instigate some discussion about how we should go about it.

To further add to what I envision, I don't really expect anything different than what people have already posted for many answers on the site (hence why I did not think it would be much more effort, and would give the authors work more exposure).

Here is a list of examples of visualizations that have been produced that I think would be wonderful additions to the blog;

I would just expect a short description, what it is supposed to convey (along with its potential usefulness or benifits over different approaches), plus some code to reproduce if possible.

share
one caveat; not to diminish the work of photographers, but snapping a photo does seem to me to be considerably easier than producing a new data visualization – Jeff Atwood♦ Aug 3 '11 at 1:12
@Jeff , I do not know (I suspect they are more comparable than you give it credit for). It doesn't have to be weekly anyway (if that is why you made the remark). Also if the graphic is produced via code it can be trivially reproduced with new data (I assume something not comparable with photographs). – Andy W Aug 3 '11 at 2:34

1 Answer

up vote 3 down vote accepted

I think it is a great idea. I'd like to add that

Further thoughts:

  • We could think of thematic issues, in the spirit of this past JSM on the Use of Graphics in Clinical Trials, where summary, opinions, or illustrations with various statistical packages would be welcome.
  • We could provide discussion about recent incorporation of new graphical displays in various software; it can be @Peter Flom's SAS PROC SGPLOT, for example, where the Scatter plot with ellipse is closely related to the question you linked to. (More information can be found on his webpage for NESUG 2011.) Likewise, the newly released Stata 12 incorporates new graphical output.
share
What specifically do you want to see from Tukey's EDA? – whuber♦ Aug 4 '11 at 12:01
@whuber The idea was merely to graphically illustrate some of the techniques discussed in Tukey's EDA or Hoaglin, Mosteller & Tukey's Understanding Robust and Exploratory Data Analysis, somewhat like a "graphical companion" to those textbooks. – chl♦ Aug 4 '11 at 12:42
OK. I am fond of both books. EDA is very graphical as it is. The real value in both, IMO, is in the analytical tools they provide for diagnosing data: determining how to re-express values, finding interactions, etc. Not all of those have made it into software even after 30 years. – whuber♦ Aug 4 '11 at 13:22
Those all sound like pretty good ideas to me. So at this point I assume it is just a matter of implementation (and getting people to volunteer). I think I would like to start out as how the photo of the week is to get people volunteering, and then let it grow into your other suggested compliments to the "pretty pictures" themselves. – Andy W Aug 4 '11 at 15:56
So, who writes the first survey post? – mbq♦ Aug 4 '11 at 18:25

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged