The whippersnapper Vikings thought their technique of riding dragons while sitting backwards rather than the conventional grabbing-it-around-the-neck method was more efficient because the dragons flew faster. But it turns out they weren’t accounting for a crucial factor — the backwards riders were picking lighter, nimbler beasts than those chosen by the older Vikings, who tended to prefer the largest, manliest creatures.
They were guilty of statistical confounding: jumping to a conclusion based on an incomplete understanding of data that has an unaccounted-for variable lurking in the background. Alan Dabney, an associate professor in the Texas A&M Department of Statistics, uses the mythical example to explain the concept to non-statisticians in his new book, The Cartoon Introduction to Statistics, due out Tuesday (July 2) and co-authored with cartoonist Grady Klein.
“The project was extremely fun for me because it was two very different perspectives trying to come together to do the same thing,” said Dabney, who joined the Texas A&M faculty in 2006. “Grady is not a statistician, and I’m certainly not an expert story-teller. He would continually ask, ‘What does probability mean in plain English, and why do I care?’ It was a challenge, but a good challenge, which is why I think it’s going to be a good book.”
Dabney envisions the book as a primer of sorts for non-mathematicians seeking to familiarize themselves with basic statistical concepts, such as confounding, probability, hypothesis testing, standard deviations and sampling distribution. He says it’s not a textbook replacement, but perhaps a supplement that will help readers develop a solid grasp of the key concepts and give students, especially non-statistics majors, a better ability to understand more in-depth details in textbooks.
A typical textbook example of confounding often taught in classrooms is concluding income bias based on gender. Dabney says data could make it inaccurately look like the salary of male graduates exceeds that of females. But what looks like gender discrimination could be more benign when a variable in addition to gender and salary — type of degree earned, for example — is factored into the equation. A science degree typically earns more than an arts degree, and men are more represented in the sciences than women, Dabney says, so looking within each degree would provide more representative data.
“To illustrate the concept of confounding, I came to Grady with the example of graduates, income and gender,” Dabney said. “His response was, ‘That’s extremely boring. What do I care about income?'”
Enter the dragons. And so the process went as the give-and-take between statistician and artist continued through their 225-page journey published by Hill and Wang, a division of New York-based Farrar, Straus and Giroux.
It wasn’t the first time that Klein, the creator of a graphic novel series called The Lost Colony, has taken a crack at illustrating an academic topic. He’s also the co-author and illustrator of The Cartoon Introduction to Economics Volume One: Microeconomics and The Cartoon Introduction to Economics Volume Two: Macroeconomics — the latter of which is described by trade magazine Publishers Weekly as “a godsend to anyone who needs a simple but complete primer on the ins and outs of economics.”
Klein and Dabney’s statistics book also has received positive preliminary reviews. Publishers Weekly called it “delightful,” while Kirkus Reviews described it as “a gentle, pleasantly illustrated induction into the strange world of bell curves and chi squares . . . A smart, enjoyable overview of this most useful branch of mathematics.”
It’s the power of statistics that Dabney primarily wanted to convey to non-mathematicians. Statistics is a tool that’s increasingly playing a greater role in the world as data becomes more abundant. In the last presidential election, statistician Nate Silver correctly predicted which way all 50 states would vote. Statisticians also are involved in cancer drug testing, space shuttle safety analysis, jury selection and just about any field in which data measurement or analysis is needed.
“I want to make it easier for non-math people to access and benefit from statistics,” Dabney said. “I think there are a lot of people in the world, not just math majors, who could greatly benefit from having basic tools for extracting confident information from data.”
Dabney’s own expertise is in the analysis and interpretation of big data, particularly that which originates from biological applications, along with non-conventional methods of teaching statistics, as evidenced by his previous project, a W.H. Freeman-produced StatClips video series.
Click here for more information about the book, including reviews and ordering specifics.
To learn more about Dabney and his teaching and research, visit http://www.stat.tamu.edu/~adabney/.
Contact: Vimal Patel, (979) 845-7246 or email@example.com or Dr. Alan Dabney, (979) 862-7581 or firstname.lastname@example.org