A study called “Booksthatmakeyoudumb” has been making the rounds, and I finally took the time to check it out. Created by one Virgil Griffith, it claims to map the correlation between one’s favorite books and one’s intelligence. According to “Booksthatmakeyoudumb,” the “smartest” book out there is Nabokov’s Lolita (it’s about white people), and the “dumbest” is Alice Walker’s The Color Purple (it’s about black people).
A closer look shows that of the “stupidest” 10 books, 8 are written by African-Americans. Naturally, this raises some serious questions about race and intelligence. It’s unfortunate that Griffith refuses to address them. But it’s unforgivable that his method actually invents them…
See, the correlation isn’t as direct as Griffith suggests. This is how it works: Griffith takes the 10 favorite books at a particular college (according to Facebook) and takes that college’s average SAT score (according to CollegeBoard). That’s it — the correlation is based on institutional and not personal preferences.
Now revisit the race issue. What kind of school would list The Color Purple as the student body’s favorite? An all-black one. And African-American institutions have a whole slew of factors — social, economic, cultural, whatever — that might impact the average SAT score of its student population. Throw in the SAT’s well-studied bias in favor of the wealthy white and you realize that “Booksthatmakeyoudumb” is just a crude repackaging the complex relationship between race and education.
At best, “Booksthatmakeyoudumb” is poorly designed. At worst, it’s just plain racist. Virgil Griffith is an irresponsible d-bag either way.
Tags: Racism, SATs, Studiesthatmakeyoudumb
An RSS feed of The Spin is available
March 5th, 2008 at 1:38 pm
This is one of the smartest things I’ve read on The Spin. There was something about the booksthatmakeyoudumb project that seemed inherently unfair and poorly reasoned to me, but I couldn’t quite put my finger on it. I think you’ve hit the nail on the head: it’s racist (or at least stupid) junk science, for all the reasons you list.
Studiesthatmakeyoudumb, indeed.
March 5th, 2008 at 2:57 pm
kudos, nick. one of the best posts i’ve ever read on the spin. more real analysis of things that matter could do wonders for this blog.
i think it’s a stretch to call the study racist on its face, but it is definitely an example of spurious correlation at its worst. a book like the bible finding its way down the bottom is another example.
even worse is saying these books “make” you dumb when they’re merely correlated with a woefully weak estimator of intelligence. there’s no causal relationship whatsoever.
the only nice thing about the graph, and something i hope is statistically significant, is how far down “i don’t read” is on the list. not reading will make you dumber.
March 10th, 2008 at 2:22 am
Despite the title, I think the website makes it clear that “correlation does not equal causation,” and it’s pretty obvious that the results are not meant to be taken seriously.
The bigger issue I take here is the inclusion of an “African American” genre at all - that’s really what makes it seem like he’s subtly trying to imply that the direct correlation that you criticize is meaningful.