I don’t think so. Virgil Griffith pulled out the top ten books read by students at various universities (it turns out Facebook collects that data for you), and then tried to correlate that with the average SAT/ACT score of each university. The result is a mess. You might be able to say that schools with low admission standards are more likely to have students who read the Bible and Fahrenheight 451, while the universities with the higher academic reputation are more likely to have students reading Lolita and Ayn Rand, but the overall distribution is more suggestive of chance — there is large, diverse pool of books read by university students, and facebook is plucking out a nearly random subset.
The display leaves a lot to be desired, too. What does the size of the lozenges mean? Standard deviation? I’d need to see something about the actual numbers for each book, too — how many universities have The Grapes of Wrath in their top ten, and how many students is the sample based on? A small college with only a few students on Facebook is a situation that is readily skewed.
I’m only mentioning this to torment you all, so you can stare at this chart trying to make sense of it as long as I did.
Gyaagh! That’s hideous. They’re saying that the “results are awesome regardless of the causation” and I’m inclined to disagree. Those results seem to be incomprehensible gibberish.
Oh, I get it! “readily skewed”. Nice!
WTF? They have Pride and Prejudice and Little Woman as “Chick Lit”, but Withuring Heights as “Classics”. Seems that we have a problem right there.
Elitist claptrap. Just about anything you read gives you more knowledge about the world, with the exception of A*** C******.
NONE of the books labeled “Philosophy” ARE PHILOSOPHY!!!!
Man that makes me angry… Life of Pi is NOT philosophy. Atlas Shrugged is NOT philosophy…
Where is Leviathan?! Where is Plato’s Republic!? Where is Descartes’ Meditations?!
THOSE are (historical) philosophy!
I guess the number of people who actually read books that are DIFFICULT is so low that they wouldn’t be considered for this chart..
One problem (aside from all the others) is that Facebook doesn’t collect the top ten most-read books at a school, it collects the top ten favorite books. So it’s not necessarily that students at various schools aren’t reading a book, it’s that they’re not liking it.
“Graphs that make you dumber”?
Graculus: I was wincing about that as well. First thing I noticed, actually.
(I’m also confused as to why the Harry Potter lozenge is so small. With the sales that thing has had, and the demographic beast that it is, I’m wondering why it doesn’t have a wider range of scores, like Dune or Ender’s Game. I wonder if there’s a social factor in being embarrassed to admit to liking something that is A) popular or B) something marketed as a kid’s book. But then I’ve never seen any embarrassed Harry Potter fans before. They’re always very enthusiastic.)
Also, I noticed, we don’t have a distribution graph showing how many respondents across scores. (But I guess that’s because he took the top ten books from a school instead of polling the people and their scores.)
I may be in a minority here, but the results do look intriguing. I’m not sure they’re valid, of course. Even if I accept his methodology and premise, the SAT itself isn’t necessarily a predictor of much, and certainly not intelligence per se. I found the chart easy to interpret, once I understood what dimensions he was trying to display: average SAT; most popular book; genre. It’s the combination of genre with their various colors that makes the chart look scattershot. Try one of the charts with links at the bottom. The “stacked” one is easier to interpret.
The width of the “lozenge” appears to relate to the width of average SAT within which that book was listed as a favorite. So, for example, “Fahrenheit 451” was a favorite in its genre only for institutions that had average SATs from about 860 to 925. The author appears to be aggregating scores across institutions, which would eliminate a lot of within-college skew. Some skew is almost inevitable, though. The Bible is probably the favorite book of seminaries and Bible colleges, which generally don’t attract the top SAT scores, for example. Some liberal arts universities have lower overall SATs because the math section lags for those students, and they may be contributing such favorites as “The Catcher in the Rye”.
I have to admit I’m astonished that some bars don’t cover more SAT scores. I would expect some to stretch for inches across the page. “Dune” and Shakespeare cover a lot of ground, but they stop around 1150. “The Hobbit” has only a small following between 1040 and 1070. “The Lord of the Rings” has even less coverage. Harry Potter barely appears at all. Perhaps college students aren’t willing to admit they like these books anymore. “Cats Cradle” came out remarkably well. However, some of the books listed, such as “Hamlet”, may well be vanity talking.
Actually,
Many of my friends consider the HP books to be crap. Generic, predictable and boring. They do love the characters and the fanfiction, though (and the general concensus seems to be that the epilogue reads like bad fanfiction. – “Leave the fanfiction to the fans, Jo.”). So it may well be that they don’t consider them among their favourites.
Nah, who am I kidding. I just have friends with good taste.
Isn’t it more like the top ten books that people want their friends to think they like? Everybody wants to be cool (at that age), and your list of favorite books is one way to signal your coolness to your peers.
Great blog!
Why does Dan Brown score so much lower than the Da Vinci Code or Angels and Demons? Or C S Lewis lower than “Mere Christianity”?
Hans – you are exactly right. Facebook is not about real life, it’s about representation. One of the few human universals is that people lie… a lot.
This chart does appear to represent something real though – it gives a fair indication of what kids in these colleges *think* qualify as good books and thus a sense of each school’s self-image and literary standards.
That said, the identities of the schools are obscured here by grouping them as SAT scores. A more interesting graph would use other axes like geography, family income differences, presence of grad programs, secular vs types of religious schools, etc.
The size of the lozenge appears to represent margin of error. Right above the chart, it says “Average SAT (with margin of error) for the 100 most popular books on facebook. ”
As I noted over on FA, it looks sort of like some of the patterns here is really reflecting what books are part of reading curricula, and what you’d have to go out of your way to read. Color Purple and Fahrenheit 451 are both common high school reading assignments. Lolita and Any Rand? Not so much.
“The Bible” and “The Holy Bible” are two separate entities, with very different numbers attached.
That should demonstrate the inanity of the exercise to anyone’s satisfaction.
It also has separate charts for “The Holy Bible” and “The Bible.”
Lolita is “erotica”? Not by any meaningful definition of erotica.
So, if some people read “The Bible” and others read “The Holy Bible”, what’s the difference?
The title to the graph is downright misleading. Causation hasn’t been shown at all. Anybody who thinks “Ender’s Game” makes you smart while “The Color Purple” makes you dumb is deeply unserious. You know, the kind of guy who thinks “Lolita” is erotica simply because sexual issues are addressed.
“Elitist claptrap. Just about anything you read gives you more knowledge about the world, with the exception of A*** C******.”
I suppose reading Ann Coulter (say her name, she poses no threat”) could teach someone that no matter how vapid and undeveloped their ideas are there is a publisher out there who will publish them.
As an aside, am I am alone in thinking Coulter maybe a transvestite with two fried eggs stuffed up his top ?
Almost no overlap between “Hamlet” and “Shakespeare”.
books that make you dumber?
forget the figures from facebook.
i recently endured the crap sandwich on paper that is thomas l. massons’, ‘why i am a spiritual vagabond’
on the recommendation of someone i expected to have better taste…(and some sense)
…painfully trite, stupid, and, like, “deep, man. really deep.”.
i’d love for pz or one of you regular commentators to shred that piece of spiritualist “work” for a while…
I notice that the Bible and the Holy Bible are different books. I wonder what the difference is? Of the two, I would think the Holy Bible would contain more truth, wisdom and knowledge but it seems to be the “makes you dumber” version by almost 150 points.
PixelFish (and others) – Harry Potter’s popularity is the reason why the lozenge is so small. The width represents the precision of the estimate of the mean, not the variation in the scores. Now the really bad news – the last person I corrected for making the same mistake was PaV at Uncommon Descent. Sorry.
I don’t think the graph is too bad, except that I’d like to see the variation in the scores – there are displays like box plots and violin plots that would work nicely. The order of the books should make some sort of sense too.
At least it wasn’t presented as a table.
Bob
I wince at every “Top 100” list that has Life of Pi rubbing shoulders with the real life-changing books.
Nice to see that Dune ranks about as high as Shakespeare.
Frankly, I can’t see how anyone would take this serious, it is a non-random sample of a completely random occurrence based on perception, taste, and peer-pressure.
However, I would be interested to see every book that the students have read that was not assigned to them, so if a student was more likely to read and enjoy Lolita voluntarily, they would be more likely to have a high SAT score, but I think that it has more to do with the students verbal SAT scores than the college.
If a certain group of people is more likely to label it “Holy”, that could explain the difference, but it would be hasty to infer that since there are so many other unexplained discrepancies in the chart.
I certainly felt dumber after reading the ‘Da Vinci Code’ (and especially after having paid for it), but I don’t see how reading unspecified Dan Brown books should be even worse.
I’ll confess, I actually have a facebook page, and there are many more sources of methodological error than already commented on. One possible source is that facebook does not distinguish graduate and undergraduate students. I attended an upper tier graduate school in my field, in a state in a different region of the country. I found after I relocated that the undergraduate program there had a poor reputation based on extremely high rates of remediation of high school English and math by freshman, grade inflation, etc. I’m sure this is not a wholly unique situation, and am fairly certain that my reading habits differed from those of the students I taught.
If these are self-reported book-readings, then the data says more about what some college students want other people to think they read, than about what they actually read, which might only be gleaned from the students’ book reports.
Humor transplants for all commenters!
These are “data” from Facebook for cryin out loud. Take it with the requisite volume of salt, get a grin out of the placement of some of your favorites, and move on with life.
p.s. The reasons for the categorizations, double-Bible listing etc. are clearly explained–check the FAQ.
It’s also pretty obvious from the chart’s very parameters that the Facebook population is itself a “bell curvish” problem. For example, it’s reasonably certain that MIT, CalTech, and Ivy-League-class schools aren’t in the sample… because the “median SAT score” doesn’t go high enough.
And notice that all of the novels are English-language books, with one startling exception? That says more than it really should.
As the name may suggest, I’m a bit saddened that Gene Wolfe’s Book of the New Sun tetralogy isn’t making more top ten lists :(
With regard to the Bible vs Holy Bible, Virgil says that it’s because of how Facebook lists them, but doesn’t Facebook list what people type as their favorite books, meaning that the only difference between the Bible and Holy Bible is what people at various schools type (leading to the FAQ’s example of Sam Houston University with both on the list of favorites)?
The fact that Fahrenheit 451 is listed on the bottom, while the Alchemist is listed near the top, is a proof that you can’t infer a correlative link between the books read and intelligence.
Why don’t you people click on a couple of the links over there before spouting off with your ill-formed opinions? The median SAT score is reported for the books, not the schools; there is a giant blue link to the list of schools included (MIT & Ivies, yep), etc. etc.
Yes, the book “data” are taken from the lists of favorite books students type into their Facebook profiles.
Interesting chart..some trends I’ve noticed (and to give myself a smattering of light justification for my opinions: I’ve worked in a major bookstore for about two years now, between and sometimes during college semesters): The ones on the lower end tend to be more ‘real life’, while the ones in the mid-higher end tend to require more imagination (or outside knowledge – eg, some idea of a setting which is outside that of everyday familiarity). Further, books that are very popular (eg, A Child Called It, The Wedding) and those that are from regular producers (eg, James Patterson) tend to be lower than those that are rarely purchased or from authors who spend more time writing (or actually WRITE) their books. And anything that’s on school reading lists gets pulled towards the middle (kids are smarter than adults, who can actually pull a book to the low end..?). Thickness seems to have less of an effect than I’d thought it would, though it definitely has one for the more massive ones, pulling them up.
So, how about we see how much the list matches up to us? Put it to the test, in a manner of speaking. Out of THAT list, my favorites* are (in the order I spotted them going down the list):
Anthem
Animal Farm
1984
A Wrinkle in Time
Night
The Hobbit
Dune
Harry Potter
Lord of the Rings
Ender’s Game
My predicted SAT score^: 1060-1098
Actual: 1580
OK – one incorrect.
*These are my favorites mostly because there were only about 3-4 others I’ve read on the list.
^Calculation method: total up the low end and high end estimates of each into separate columns, then divide each total by ten to make the range.
(btw, I do know it’s not serious. Still fun to play with though!)
1. Not everyone at a school is on facebook.
2. Not everyone on facebook lists their favorite books.
3. As everyone has mentioned already, the books on a person’s list might not be one’s actual favorites but what one wants to project as being one’s favorites.
4. SAT is a poor predictor of intelligence.
The data gathering method is ridiculous! This whole thing is silly and the more I think about it the more I want to not think about it anymore.
C.E. Petit:
“It’s also pretty obvious from the chart’s very parameters that the Facebook population is itself a “bell curvish” problem. For example, it’s reasonably certain that MIT, CalTech, and Ivy-League-class schools aren’t in the sample… because the “median SAT score” doesn’t go high enough.”
Those schools certainly are. The guy who made this is a Caltech grad student and if you look at the page with the schools included Caltech’s the first one listed, but none of the books listed seem to be exclusively read at these schools.
Spinoza et al.
I can’t claim to get the categories either. At best it can be argued that some of the philosophy ones have philosophical undertones (despite being fiction stories). It may not be very insightful or sensible philosophy, but there’s a lot of stuff but there’s a lot of stuff that’s neither that seems to get that title.
And yeah, samples from a site like Facebook will surely have some weird biases. It’s not clear how representative a cross-section people who sign up and decide to list their favorite books would be.
Nitpick: Fahrenheit, Fahrenheit 451 – or 500 K, for us furriners.
Nitpick: Fahrenheit, Fahrenheit 451 – or 500 K, for us furriners.
dreikin:
Since the mean score estimates only go up to about 1350, and since my SAT score was also a couple hundred points higher than that, I can tell you without going to the trouble of enumerating a list of favorites that my actual score also falls well outside the predicted range.
WHAT??? Lolita is classed as erotica?
That is completely fucking INSANE. Clearly, the authors of this chart didn’t go to one of them fancy high-SAT schools that actually READ that book.
I can assure Ian, dreikin, and all potential commenters that, no matter how Mensa-like your abilities, nobody gives a shit about your SAT scores.
True, there’s no REAL need for numbers (and we can’t confirm them _anyway_), but how else can we get a good inaccurate estimate of the accuracy of the chart?
Yes, Sven, I know nobody cares. Bragging over SAT scores was not the point (I not very fond of most of these type of standardized tests in the first place)
Humor transplant for Sven? What fun are silly internet studies if you aren’t allowed to overanalyze them on Pharyngula?
hmmm…perhaps I am in a foul mood today. Apologies, all. Must of been that potato salad…
yeah…must’ve been…
It occurs to me that instead of comparing the books we read to our own SAT scores, we should compare them to our schools’ average SAT scores.
From the list of schools at that site, my college’s average SAT score was 1219.
My favorite books from that list are, in no particular order:
Wicked
The Chronicles of Narnia
The Lord of the Rings
Dune
Harry Potter
Alice in Wonderland
A Wrinkle in Time
Ender’s Game
Cat’s Cradle
Lolita
(Before I get criticized for including the top-ranked book from that site, let me just note that Lolita really does happen to be one of my favorites.)
From those, my school’s predicted mean SAT score falls in the range 1094.25 – 1142.35. Swing and a miss.
From Facebook, my school’s favorite books are:
Lord of the Rings
Harry Potter
Grapes of Wrath
Fight Club
Catcher in the Rye
1984
(and four others that didn’t make the list)
That gives a range of 1068.85 – 1096.15. Bzzt, wrong again.
Curse you PZ, I fell for your trap and spent way too much time thinking about that graph. It was a stunningly useless way to represent the “data” he collected, and I’ll cut my losses by not writing out why.
When I saw the post title, I immediately thought “Anything written by the DI or Creationists”.
I still think I am correct.
I have no idea what a SAT score is but I’m feeling a bit smart arsey today so I’m going to award myself a 1350. However the only thing I have read today is the instruction leaflet from inside a tampon packet. That’ll fuck with their correlations.
Wait, Atlas Shrugged isn’t in the Dystopian category. Atlas fucking Shrugged? It’s a book about civilization collapsing while a handful of elites make smores on the funeral pyre of the world. How much more dystopian can you get?
Might the Bible vs. Holy Bible issue have to do with there being a difference between the Bible Catholics use and the Bible that Protestants use?
Leaving aside the sampling and representation issues, what stood out to me is all the *assigned reading* (34 out of 100 books). I was going to say students don’t read (mutter, mutter, kids these days…) and haven’t read any other books.
But then I looked a few dozen profiles and realized that the students read a lot. So now I think there is likely a bias in that school assigned reading overlaps a lot while other books read are more dispersed. And the list is likely just noise.
There is definitely something fishy about that data.
Lolita is erotica and Atlas Shrugged is philosophy.
I just threw up in my mouth a little.
PZ Myers,
The graph is adequately explained, but you must click the “[ BROWSE BOOKS ]” or “[ BROWSE SCHOOLS ]” buttons at the top of the page. You can see complete details on the distribution, error bars, and even a histogram for how the average of each book is calculated.
Furthermore, your criticisms of “I’d need to see something about the actual numbers for each book, too — how many universities have The Grapes of Wrath in their top ten, and how many students is the sample based on? A small college with only a few students on Facebook is a situation that is readily skewed.” are fully explained on the Bookdetails for every book.
Here’s the Grapes of Wrath:
http://booksthatmakeyoudumb.virgil.gr/bookdetails.php?book=The+Grapes+Of+Wrath
-Virgil
Virgil,
WTF is a “Rank Within Religion”?. Comforting at least to see that the “Holy Bible” made it at 10 out of 10 on that scale. Not that it was a articially imposed value or anything. I’d never suggest that.
Um,
That should be artificially. Still, the mind boggles as to what else was ever going to get a high value in the “Rank Within Religion” category.
Virgil, I’m calling a huge BS on this and I think you know it.
To remain forever dumb, only one book is required as a
constant reader: the bible. All else is moot.
These are “data” from Facebook for cryin out loud. Take it with the requisite volume of salt, get a grin out of the placement of some of your favorites, and move on with life.
And tequila?
Back on topic, I agree with the various posters that pointed out that anything one puts on Facebook is how one wants to be perceived by their friends. There is no doubt a lot of posturing and lying going on here. I would love to assess my reading preferences with reference to my SAT score but a non-American education means I can’t. Boo hoo :o(
C’mon you godless heathens… I’m atheist and even I know that the reason there are two bibles is because one is Kjv and the other is NIV. [you’re a heretic if you pick the wrong one.]
(Posted from my couch with a Wii! A first!)
Some of you might not have realized, that the name ‘booksthatmakeyoudumb’is simply a name — if you’d read later on, you would’ve read that yes, correlation does not equal causation.
While I do agree that there might be some categorical, descriptive, and aesthetic work left to be done, the idea behind this is interestingly valid. Correlation studies are ubiquitous and extremely useful in every area of science. By themselves, they show interesting facts about the population.
I could go on a lot more about the validity of correlation studies, but I thought I would just point out the basics…for those of you who discount a study as “gibberish” just because you personally don’t find it interesting.
i like this graph because i’ve read lolita and 100 years of solitude. i are smart.
Oh I remember that kind of graph from my stat class.
It’s called a ‘Crock of shit’ graph.
oooh, must read Lolita, it sounds dirty. I’m reading DH Lawrence at the moment and, quite frankly, he’s disapointingly unfilthy.
This is stupid. I feel dumber for looking at this graph.
_Their Eyes Were Watching God_ and _The Color Purple_ are both good books, and the former is one of my favorites, with its lyrical prose and earthy plot.
And you know what? My dumb SAT score registers to the far right.
Hey, I’m just celebrating that many college students HAVE favorite books (or say they do). And that Shakespeare made the list!
Hmmmm . . . I spend my days finding invariants using the De Rham cohomology, teaching students basic calc, etc.
At night I like curling up in bed with Cheetos and my dogs and reading detective stories about Hard Boiled Dicks – Philip Marlowe and the like. I don’t think anyone would accuse me of being dumb, but I don’t think Raymond Chandler wrote what anyone would call deathless prose, either.
In fact, that last ‘hard’ book I read purely for entertainment was, I think, “The Ancestors Tale”, and that was right after Christmas break started. Anybody here read Thucydides in bed to relax after a long day? Or a little Descarte, “Truth in Science”, say?
I think what you’re seeing here is Rex Stout for relaxation, not Alice Walker for serious[1].
[1]Who, btw, is highly over-rated, mostly for political reasons. Read Eudora Welty instead, if you must.
Whoa, whoa, whoa.
So, um…
…what you’re saying is…
The data is invalid?