Jan 25

Artur Bergman

Artur Bergman

Books that make you dumb

Wikiscanner hacker Virgil Griffth told me a while ago about his latest data mining project, to visualise the relationship between books and SAT scores. Today he released his findings at Booksthatmakeyoudumb.

He does this by cross referencing the 10 most popular books at every college, as given by Facebook, and the average SAT score. He then presents it all in this nifty little visualisation.

I find it somewhat amusing and surprising that erotica takes top and bottom positions, with Lolita at the top and the author Zane coming in last (perhaps it says something that the lowest scoring book is actually miscategorized.) The book named "I don't read" also comes pretty far down.

In all, the results aren't that surprising, but as Virgil said to me; "but isn't it wonderful to have concrete data to back it up?"

tags:   | comments: 36   | Sphere It

Previous  |  Next

0 TrackBacks

TrackBack URL for this entry:

Comments: 36

  Fnarf [01.25.08 11:43 AM]

Oh, fer chrissakes, "Lolita" isn't erotica. It's the best and smartest novel ever written by an American, is all.

  MH [01.25.08 11:50 AM]

"Nabokov" is a mighty interesting name for an American...

  Adam [01.25.08 11:52 AM]

Nabokov was not an American.

  Bottle Imp [01.25.08 11:54 AM]

I'm with Fnarf about "Lolita" not being erotica... but he was Russian.

  Nicolas C. [01.25.08 11:57 AM]

Aren't we mixing cause and consequence here?
Shouldn't this be called "what do dumb people read?"

  theophylact [01.25.08 12:04 PM]

"Written in English", perhaps. He did become a naturalized US citizen in 1945.

  Alek [01.25.08 12:17 PM]

So the conclusion is that people with over 1350 do not read books?

  Martin [01.25.08 12:30 PM]

Looks more like the people who read more than their English Lit class required do better on exams. Duh?

  Kristina [01.25.08 12:35 PM]

Iiinteresting, but not a very accurate way of conducting a study. Pretty sure everyone is not listing their favorite books on Facebook...and of those that are, how can you be sure where their SAT scores fall relative to the average at their school? I think Nicolas C. is on to something, as well.

  Misread [01.25.08 12:38 PM]

"but isn't it wonderful to have concrete data to back it up?"

Facebook? Really? Granted, it is an interesting project.

  roman [01.25.08 12:39 PM]

I completely agree with the "cause and concequence" critisism. I think the chart is nice, but about as far from meaningful as possible.
Since we're data mining anyway, why not compare the average height at colleges to SAT scores. Or the average number of friends they have.
Also, the sat doesn't measure how smart you are, it measures your vocabulary. Maybe the graph can be titled, "Books that use the same words that are found on the SAT"

  Charles Jolley [01.25.08 12:41 PM]

This is a horrible use of data and I'm surprised you would propagate it on a blog like this. Even if the books people read were correlated with their SAT scores there are almost certainly other factors at play here. For example:

1. Look at the average lower SAT scores of African-American literature. The SATs generally do a poor job of measuring the intelligence of African-Americans (who likely comprise a larger segment of the population reading these books).

2. The Bible is read by such a broad audience that it probably correlates weakly to any measure of intelligence compared to a book like Freakonomics would (which is more targeted at the general demographic that does well on the SAT).

A much better measure would have been to put together a graph showing the correlation of various books to SAT scores. That will would not establish causality but at least you could draw some statistically relevant conclusions from it.

The only thing that might be useful about this chart is that it demonstrates how amazingly easy it is to misuse the abundance of data we have available on the net to wrongly draw conclusions or confirm predispositions.

  Ms. Blue Sky [01.25.08 12:42 PM]

I believe Fnarf was making a joke.

  wiredDice [01.25.08 12:47 PM]

Interesting list.

I like the fact that I've read over 90% of those and somehow managed to get a 1570 on my SATs.

Guess I must have cheated somehow...but then the missing 30 points were from the English section.

  Alex H. [01.25.08 01:07 PM]

Agree about Lolita: it's only erotica if you haven't actually read it. Yes, it contains sex, as do many of these (cf, Choke).

The real question is whether the average university SAT score is represented in active users of Facebook. Not casting aspersions, it's just an open question.

  Sasha [01.25.08 01:18 PM]

Re: Nabokov.

He was naturalized in 1945 and wrote Lolita in 1955, so I would say yes, it is a novel by an American. I suppose it's possible that he wrote parts of it before he was naturalized, but that seems unlikely, no?

  gene [01.25.08 01:34 PM]

Aw man, a Dan Brown book is on the upper end of the SAT scores? Really? We're in trouble. Start stockpiling water and cans of food NOW

  Andrew Larcombe [01.25.08 01:57 PM]

Blogs that make you dumb, more like.

Can't believe I wasted a mouse click on this tripe.

  Bob [01.25.08 02:56 PM]

And I can't believe you wasted more time typing that you can't believe you wasted time.

  Mitch [01.25.08 04:08 PM]

So what is the difference between The Bible and The Holy Bible? Big difference in the SAT scores.

  Fnarf [01.25.08 05:15 PM]

"Adam" is a pretty interesting name for an American, too, if that's your standard. Nabokov lived in the US for decades and was a naturalized citizen. That makes him an American, from where I'm sitting. And "Lolita" is a deeply American book, saturated with our culture and landscape. It is the absolute American masterpiece. Nabokov learned to write in English before Russian, by the way; he grew up trilingual. "Lolita" is his greatest book, and the greatest American novel ever written. To classify it as erotica is illiterate. It is profoundly sad.

  Peter [01.25.08 09:37 PM]

Lolita is not even close to erotica.

  Billy-Bob [01.26.08 03:04 AM]

"I choose to conclude that people with over 1350 don't waste there time on facebook."

(Astute observation, but it's their time, not "there" time).

Props to all who GOT THE JOKE. the rest of y'all need some serious drugs or some serious way to lose your seriousness. seriously. ok, now i'm annoying myself.

  Tom [01.26.08 07:27 AM]

The author should read Freakonomics...

These aren't causal relationships, but rather correlated.

The analogy that I find helpful in describing causal vs. correlation (taken from Freakonomics) is the relationship of successful people to homes that provided books during a person's childhood.

The mere existence of a book in a home doesn't make one successful. Rather, parents that encourage reading are typically better educated, higher income, etc. and apt to provide more opportunities for a child.

  Mo [01.26.08 11:47 AM]

Holy carp. There's a school that a) isn't a branch of Brigham Young and b) isn't in Utah that totally loves The Book of Mormon?

  aasdf [01.26.08 05:40 PM]

the person who made this is engaging in a classist, racist and sexist measure of intelligence. . the sat is tailored to rich white men who grow up in the suburbs. rich kids who dont have to work a job to go to school have more time to expand their reading. you think someone who works 40 hours a week just to go to class for 16 credits and eat ramen noodles and hopefully sleep has time to spend reading something like lolita or even reading at all if it is not homework? the assumption that reading is an indicator of intelligence (as is assumed by the creator of the survey on their webpage) is absurd.

  ecological fallacy [01.26.08 10:17 PM]

You can only make a meaningful correlation of this kind at an individual scale (with individual's score and favorite books). Using college-wide data means your method is total non-sense. It is like trying to conclude that "black people are less educated" with a correlation between large black population and low literacy rate at the state level. it is simply the wrong scale of measurement.

  Tim O'Reilly [01.27.08 12:24 AM]

ecological fallacy (and others) --

Clearly, Virgil's title "Books that make you dumb" is inappropriate and misleading.

But it's interesting to realize that almost all medical studies suffer from this same fallacy. One of the interesting things I learned talking with some geneticists at Davos is that now that they realize that there is huge genetic variation in response to drugs (e.g. some people metabolize caffeine at 5x the rate of others, and thus can drink much more coffee without getting any buzz), you can see the flaw in many medical studies.

You might for example have a drug that increases survival by 85% for 30% of the population, and does nothing for the other 70%. It will show up in blended research results as much less effective than it actually is for the people it does work for.

This is why personal genomics is going to be so transformative a technology.

  Maciej Ceglowski [01.27.08 03:10 PM]

John Updike solved the problem of assigning a nationality to Nabokov by describing him as something like "the best author currently writing in the English language". Calling Nabokov an American says more about the limits of such labels than about the writer.

And If you think Lolita is erotica then you are reading it wrong!

  Joe Cardali [01.28.08 05:19 AM]

I'm surprised to see actual good books such as Dune and Ender's Game up there...

  chrisco [10.27.08 12:58 PM]

Seems like a fun exercise with a fun title, which itself serves a very humorous purpose, assuming Virgil chose it on purpose, specifically because it was a logical fallacy. In fact, I'm chuckling (on the inside) right now the same way I did when a nit-wit in class raised his hand to answer an easy question that smarter kids passed on by choice. It's just plain funny.

  Dr. Funke [12.12.08 08:25 AM]

Wow, people need to chill out on the whole correlation/causation thing. Yes, the books in this study do not necessarily make you dumb, and there's plenty of good books at the lower half of the scale. What it does show is how few people read outside of school. I suspect some of the books are also what students THINK should be their favorite books.

  eliz [03.01.09 09:27 PM]

okaaaaaaay-- let's see. whatever you interpret "intelligence" to be will determine how offended or embraced you feel when reading this chart. clearly, those who spend time perusing the works deemed by those who author the SAT to be "important" and "enriching" are going to do better on said test, n'est-ce pas? study after study has shown listening to classical music, with it's often complex and always metered rhythmical structure and tonal and counter tonal system of sound increases brainwave activity and expands the conscious mind. also, isn't it just plain common sense? a slim majority of the population listens to beethoven, a slim majority of the population has above-average intelligence. above average intelligence, by it's very definition, leaves a broad many out in the cold, as does the harvard admissions dept. and the SAT. and that's the way it is. too bad, so sad, little wayne and stephenie myers fans.

  J. [03.01.09 09:31 PM]

Yep, gotta love what a genius one can become by reading overtly sexist crap about silly oversexualized little girls being pined over by neurotic old men.

Also unsurprisingly, Zane is a black woman whose erotica is appreciated in the feminist community. And, gee, coincidentally, most of the "low-scoring" artists are black! Surely we couldn't get someone to point out the total sociological ignorance of this "list."

  Jon R [03.02.09 07:52 AM]

If you think Lolita was erotic, then there's something terribly, terribly wrong with you.

  Anonymous [04.04.09 09:08 AM]

Really? I'm surprised that people half smart would read "The Alchemist". Only if the translator was way too good. I just read it in Portuguese. It would make sense to me to find it near "The Da Vinci Code".

Post A Comment:

 (please be patient, comments may take awhile to post)

Type the characters you see in the picture above.