Yes, I’m a little bit late getting to this one, but to be fair, I was in Seattle when the 2011 Survey came out. In fact, I probably never would have come across it had someone not noted my interest in Joshua Goldman (re: Joshua Goldman’s guest column in the DP, “My ‘Birthwrong’ experience”) and forwarded me this sample of his other work.
To be clear, I’m not here to pick on Joshua Goldman – I’m here to pick on whoever came up with these (though ‘whoever came up with these’ could very easily be Joshua Goldman):
I trust that I don’t have to explain what’s wrong with what you just saw (hint: it’s not the topic, or even the content), but if you can’t figure it out, ask the most nonjudgmental friend you have (i.e. not me) for some more help. I think my favorite part is that the filename for that first graph is ‘sex-toys-try2.jpg’, which means someone made it, realized he fucked up, went back and fixed it, and still ended up creating the Josh Goldman column* of graphic design.
*OK, I guess this is here to pick on Joshua Goldman.
Finally, in the spirit of the Beacon’s latest viral hit, I wanted to share one more survey tidbit:
I feel like those numbers could have been ordered better to maximize impact.
And that makes three posts in a row I’ve tagged ‘sex’.
Paper Treiger 
Judge away, but I can’t find anything wrong.
“Sex-toys-try2” is not a second try, rather what sex toys have people “tried.” No one versions things as “try2.” The final category of that chart demonstrates the the question had the word “try” or “tried” in it.
The first chart can be wrong if the sum of “never tried” and the max value of any of the other categories is not 100%, because those are mutually exclusive–it seems to pass that test.
The second chart can be wrong if all the %s don’t add up to 100%, because each category is mutually exclusive. I didn’t sit there with a magnifying glass, but it looks right.
Spelling looks right, lower case appears to be a stylic choice, which is pretty popular these days.
i would have preferred if the axes had units, like (%), but it’s pretty obvious.
LikeLike
Poor Daniel S,
I will give you a hint as to what is bothering Mr. PaperTreiger:
Data in the form of lines! Let’s calculate their slopes! And use them to interpret and predict things!
LikeLike
At one point I suspected that’s what was bothering you, but I figured you probably realized why he used lines.
The lines allow him to compare between responses AND between classes. Imagine how difficult it would be to make those comparison without the lines, or most any other type of graph. There are very few graphs that let you do that very well, and none as immediately intuitive as a line graph.
“Spider charts” are a well-known example of using lines not for trend, rather to improve ease of comparison for two types of date.
Checkmate.
LikeLike
I think a bar graph would have worked just fine.
LikeLike
Thanks for picking on me! No, I didn’t make those graphs. We have a graphics person/editor at the magazine. I supervised the survey and wrote that cute little blurb at the beginning. We did not claim, and still maintain, that this was not a scientific survey. We didn’t have the resources, nor the expertise. Please just take it for what it’s worth.
LikeLike
*correction: We did not claim that this was a scientific survey, and still maintain that. Sorry for my bad wording. Here was our sampling: “We sent out a comprehensive survey to a random sample of 1,500 anonymous undergraduates and received over 500 responses.” Clearly there is a response bias and a whole load of other factors going on. So, as I said, take it for what it’s worth.
LikeLike
Joshua, Thanks for stopping by – but whether the survey was scientific or not was not really what I was picking on 🙂 Scroll down for the trackback article, where I do my best to clarify
LikeLike