Stat content in XKCD webcomic

By Douglas Andrews posted 09-09-2012 16:12

Recommend

Over the past summer I gradually picked through back issues of the webcomic XKCD at

http://www.xkcd.org/

to find some with stat content for use in my classes. If you’re not familiar with this strip, the author characterizes it as “a webcomic of romance, sarcasm, math, and language” and issues the following warning: “this comic occasionally contains strong language (which may be unsuitable for children), unusual humor (which may be unsuitable for adults), and advanced mathematics (which may be unsuitable for liberal-arts majors).” Overall it can be a bit crude or cynical or sexual or juvenile or meta for some tastes, but there’s nothing there that would make modern students blanch. Well, I guess that’s not saying much.

The author is up to about 1100 comics at this point. And while a plurality of them seem to depend on specific knowledge of comp-related nerdy stuff for their humor (which I often don’t get), many turn more broadly around science stuff, and I found a couple dozen that have at least some connection with statistics or probability. I’ve listed them below. If you know of others, please let me know. There are many others in the strip that pertain to math, but I haven't catalogued those. And there a couple that are a bit too crude which I have omitted from this posting.

Bill Huber notes that many of the comics linked below are already part of a "What's your favorite data analysis cartoon?" list at

http://stats.stackexchange.com/questions/423/what-is-your-favorite-data-analysis-cartoon

which is part of the "Stack Exchange" site, billed as "a collaboratively edited question and answer site for statisticians, data analysts, data miners and data visualization experts."

There are only a few that scream to me for immediate stat ed use:

http://www.xkcd.org/552/ “Correlation” drives at the falacy that association implies causation, though in a subtle way that might take some explaining for folks.

http://www.xkcd.org/882/ “Significant” is great for communicating the dangers of false positive results from multiple analyses.

http://www.xkcd.org/892/ “Null Hypothesis” is a bit more overtly meta than “Correlation”.

The rest are less directly connected – often just mentioning a particular probability distribution, or the census bureau, etc. There are several that feature some kind of stat graph, like a bar graph or a time series plot; I didn’t list most of these, but included a few that had some additional kind of stat or prob content.

http://www.xkcd.org/12/ “Poisson” is a bit lame, but it’s still about the poisson distribution.

http://www.xkcd.org/30/ “Donner” is a hoot for those who use the Donner Party data.

http://www.xkcd.org/190/ “IPoD”

http://www.xkcd.org/221/ “Random Number”

http://www.xkcd.org/314/ “Dating Pools”

http://www.xkcd.org/369/ “Dangers”

http://www.xkcd.org/373/ “The Data So Far”

http://www.xkcd.org/493/ “Actuarial”

http://www.xkcd.org/523/ “Decline” is a warning for data freaks, but is also about correlation/causation again.

http://www.xkcd.org/539/ “Boyfriend” shows an outlier on a boxplot.

http://www.xkcd.org/563/ “Fermirotica”

http://www.xkcd.org/605/ “Extrapolating”