I’ve been enjoying William Cleveland’s The Elements of Graphing Data, a book I wish I’d discovered years ago. The following sentence jumped out at me:
No complete prescription can be designed to allow us to proceed mechanically and to relieve us of thinking hard. (p. 59)
The context was — well, it doesn’t matter what the context was. It’s a great encapsulation of what statistical teaching, mentoring, and consulting should be (teaching how to think hard) and cannot be (mechanical prescriptions).
In an exercise described as “rigorous mapping of ridiculous data,” Kansas State geography student Thomas Vought plotted the geographic distribution of the 7 deadly sins for a poster presented at the Association of American Geographers conference.
Many of the maps aren’t very kind to the Deep South. I was somewhat disappointed to see that my county is fairly nondescript — neither sinful nor virtuous — on 6 of 7 indices. But we are apparently quite the hotspot for envy.
The ridiculousness isn’t so much the data itself as the interpretations (which I’m sure Vought wasn’t entirely serious about). Lust, for example, is indexed by STDs per capita. That doesn’t necessarily mean that you’re having more sex with more partners — just that you’re not being very careful about it.
My region’s supposed sin of choice, envy, is indexed by thefts (burglary, robbery, etc.). I doubt that most of those crimes are really about envy. My bike was stolen last fall, but odds are the thief wasn’t coveting the bike itself. They probably just fenced it for some meth.
The conference location, Las Vegas, probably helped motivate Vought’s whimsical presentation. My main conference will be in Vegas next year. Maybe I should think about a followup?
Paul Krugman writes about Robert Rubin, Alan Greenspan, and Lawrence Summers, who appeared on the cover of Time in 1999:
Two… have since succumbed to the magazine cover curse, the plunge in reputation that so often follows lionization in the media.
Of course, following best practices in data visualization would mean you should show the central tendency and the variability (in all directions). I’m not holding my breath for density plots on the nightly news, though. A single, typical exemplar would still be an improvement over a single, cherrypicked extreme.
However… Maybe I’m too jaded, but I wonder about unintended consequences. For example, will there be a flood of crappy research after this rule? If companies are required to depict “typical” results, they may churn out poorly designed studies to get the numbers they want, hoping to lend more credibility to bogus products. And if these studies are marketed as “scientific” and then easily (and publicly) disputed, that could feed into kneejerk cynicism in the public about science more broadly.
Consider that FDA clinical trials are one of the most highly regulated forms of research around, with numerous checks and balances designed to ensure integrity. The system mostly works, but there are still serious concerns about conflicts of interest. How well is the FTC going to ensure the quality of research on consumer products, herbal supplements, diet plans, and the like? Will there be independent investigators, peer review, mandatory publication of negative results, etc.?