December 16, 2010
Does everybody secretly hate graphs?
I'm reading the book Fault Lines by former IMF chief economist Raghuram G. Rajan of the U of Chicago economics dept. It's a pretty good read, but what struck me while flipping through it is that it's solid text: no quantitative graphics, no tables of numbers, just paragraphs. The sole thing to interrupt the flow of paragraphs across a couple of hundred pages is a poem by 18th century economist Bernard de Mandeville.
That lack of tables and graphs can't be natural for an economics professor, can it? The publisher must have told him what statistician Andrew Gelman's publisher told him when he wrote Red State, Blue State, that each graph in the book cuts sales in half.
So, does everybody really hate graphs? If so, why does everybody who gives a presentation think they have to do it in Powerpoint? (As in Abe Lincoln's Gettysburg Address graph comparing New Nations -87 Years to Now.) Do audience members actually look at the graphs? Or do they just appreciate the chance during the work day to veg out in a dark room with a glowing screen? "Powerpoint: It's Almost Like Watching TV While Getting Paid!"
Or did people used to like graphs until Powerpoint came along?
Do the books that the people who sit next to you on the airplane read have graphs in them? A large fraction of people in airplanes are traveling to meetings infested by Powerpoint graphs: if there is a graph in their book, do they consider it work?
I like graphs. I had Minard's now-famous graph-map of Napoleon's 1812 invasion of Russia up on my office wall for years. It worked as a sort of secret club handshake for people walking by. One out every zillion people who walked down the hall past my office would recognize it and introduce himself.
I find that graphs always take me about five times longer to finish creating than I expected. For example, in VDARE, I've got a graph that answers the obvious question (obvious to you and me, at least) about those PISA international school achievement test scores that everybody was pontificating upon last week. My graph shows where the PISA test scores of the four main racial/ethnic groups in the U.S. would fall compared to the national average scores for the 65 countries that took the test.
A simple idea, right? Yet it took me forever to get the graph right so the answer is clear. (First, I had to find the American race numbers, which don't appear in any of the hundreds of pages of data posted by the OECD last week.) Then, I had to fiddle for hours to get the graph to look less confusing.
I think it finally came out okay. But, if everybody out there secretly hates graphs, then I'll stop wasting time making them.
Subscribe to:
Post Comments (Atom)
82 comments:
It's easier to make the data fit your argument if you leave out the data that doesn't agree. This is harder to do in a graph than it is in paragraph format.
Graphs aren't found in popular books, probably because they distrupt the flow of reading too much.
I think the easiest thing to read are short, concise paragraphs with large print.
If you have to use a graph to get your point accross, then the idea is probably too complicated for the unwashed masses to read for pleasure.
Graphs = technical presentations.
As a physicist, almost all the presentations I see or make are 95% plots and tables. Even equations are too much text. When I write papers and reports, I make the plots firsts and then write the text. Physicists are just plot people, I guess.
Among those "real" people I know who read my blog, they are unanimous in never reading any post with graphs - they like it when I just rant, whether they agree with me or not. I much prefer to see real facts in posts, personally.
And unless your blogging name is "Razib", it definitely is a major time investment to produce any kinds of graphs or charts.
I love graphs. I read some blog posts by Rajan a while back, and liked his thinking (which overlaps with Steve's ideas), but now that you say that his book has no graphs in it, it actually makes me less likely to buy it...
I like graphs. They let you understand facts and their relationships much faster than text.
Of course, if they are used inappropriately, they can be useless or annoying. But you can say that about all useful things.
My Hansonian hypothesis is that in the case of tables, it is not expected that the reader actually study them. They are there simply as a certificate of credibility for what the author has to say about the subject. On the other hand, a good graph invites you to study it, which authors only want to happen if they care about conveying information more than seeming credible and getting published.
I like graphs very much.
Fred N.
Your VDARE link leads only to a solicitation for money. Please post a link to your actual graph, or better yet post it on your blog.
I am more likely to buy a book with graphs.
I do not claim to represent the general public, but I love graphs. Text has nothing like their ability to instantly convey not only huge amounts of info but also concrete relationships between many variables.
Even if you wrote all the info down in text, it still wouldn't work as well because people couldn't plot all the stuff in their minds and "see" the relationships. And it would be incredibly unpleasant reading.
I'd always assumed that the general scarcity of graphs reflected printing costs -- particularly the very high cost of color printing, which can convey huge amounts of info via very easy-to-consume graphs. One of the main reasons I was excited about the switch to ebooks was that it would eliminate the extra costs of charts.
(Yes, there'd still be the cost of actually gathering the data and plotting it, which is harder than it seems. I worked as a newspaper reporter and editors always thought it possible to bang out a good chart in about five minutes when in reality it took hours to put together anything that would actually enhance your story.)
I'd be really interested to know the thinking behind the "no charts" idea in the realm of serious non-fiction. And I think a lot of your readership would as well because I'm guessing you've got a lot of chart fans here. Really I can't see how you could write serious non-fiction in many areas without massive graph usage.
I am a data analyst and my boss wants graphs! He specifically wants the kind of graphs that are found in The Economist. He was also talking about bringing in a graph expert to train the staff on how to build graphs. You would think that the way our society is today, that graph would have the perfect condensed visual picture that people wopuld be looking for. Reading is a lot of work!
I like graphs, I have the Napoleon's invasion one (a copy from Tufte) and more to the point I think the reason publishers hate them is fear that someone will find them politically incorrect. Soft major types can never quite extract numbers from prose.
Graphs are an excellent tool to the extent they help convey complicated results in an easy-to-perceive format. All too often, especially in the business world, they serve no useful purpose other than looking nice and making the presentation seem more authoritative.
Peter
long life to graphs and to effervescent Mr. Sailer !
I love graphs. All smart numerate people love graphs. They are well worth the trouble to make. Making graphs only loses you the morons, and let's face it Steve, you have never had much chance of becoming very popular with morons.
I like elegant, clear, clever graphs. I dislike fusspot, overcrowded, chumpish graphs.
P.S. And I can't stand bloody pie charts. Stupid things.
Hate graphs?! Love graphs.
Love the graphs.
I love graphs and/or pictures! They're a great way to break up a book, and I am a visual learner.
I prefer graphs(and illustrations). They usually communicate an idea better than walls of text do.
I love graphs, and usually click to see them (which I can't do for yours, at the moment, because VDare is blocked for fund-raising). I had saved an ad showing the Minard Napoleon graph, and when Edward Tufte brought his all-day presentation on information display to Denver I went -- it was spectacular, and no Power Point at all except as examples of what not to do.
I love graphs, but then again I have an unhealthy attraction to statistics.
I like graphs, but only if all the information needed to interpret the graph are contained in the graph itself. I frequently see online graphs that don't label or define the axes, or leave the numbers off, or maybe put the maximum on but fail to indicate where the origin is, or in other ways leave you in the dark what the scaling is.
I think it was in "How to Succeed in Busines Without Really Trying" where the protagonist simply flips a graph on its side to change a down trend into an up trend.
Well, not really. For economics they make sense, but I hated making graphs in functional analysis, and any other math.
Oh, and I hate the whole Powerpoint craze too. The overhead projector, the lights turned down, the speaker mindlessly following along and reading it out loud in your run-of-the-presentations. And they give you a handout and it's exactly the same thing, only on paper. So you read along..gah.
Graphs are great. You have a pretty smart audience, it's not like graphs are lost on us. I don't think anyone here reads Tom Friedman books.
Perhaps people don't like graphs but obviously people need graphs. For example, the big debate in congress toady is the fight over the so called Bush Tax Cuts. The Democrats aver that if you don't collect 100 billion dollars you will be short 100 billion dollars. That's an easy enough idea that you don't need a graph or even an explanation.
However the Republicans try to respond with the "Laffer Curve" but they never show the curve - must be that popular loathing for graphs. In fact the argument for higher revenues from lower rates did not originate with Laffer. Melon said much the same thing in the nineteen twenties and Hamilton in the seventeen nineties.
The concept behind the argument for lower tax rates is based on elasticity. If government was unconstrained by any other forces it would maximize its revenues by setting its tax rates at the point of unitary elasticity - a point on a curve. You can explain this concept in plain text but it is inherently a graphical concept.
The Republicans seem to fear graphs on this issue. The result of which is that the public wanders about in a fog of charges and counter-charges.
The only graph routinely shown on TV news is the graph of Obama's declining popularity.
The most famous political graph of recent history was the infamous "Hockey Stick" graph of Michael Mann. The unsettling aspect of that controversy was that the manipulation by Mann was so simple and so obviously wrong. Yet it was almost never explained in simple terms.
I don't think it's graphical representation per se. I think it's any quantitative idea. When a prospective buyer thumbs through a book in a book store and he sees graphs it is a warning sign - quantitative notions lie within!
Albertosaurus
So this post is just a scam to send people to some VDARE donation site? Pretty cheesy, actually. I was looking forward to seeing the graph that you spent so much time on, but then I wound up feeling phished...
The ability to write clear text is a necessary but insufficient condition for being able to make a clear graph or schematic. If a graph is confusing, it probably has too much detail or is being used in aid of a muddled explanation.
For decades The Scientific American set the standard for the graph makers art. The entire content of any feature article could be gleaned from the figures and graphs. To this day I remember one figure showing the density of nerve endings in the human body represented by a severely distorted homunculus, where the dimensions of the body parts were scaled to reflect the nerve density. The result was a creature that looked like the dimwitted negro hunter in the banned Bugs Bunny cartoon "All this and Rabbit Stew."
I think the 5 to 10% of the population in quantitative professions (e.g., engineers, financial analysts, accountants) mostly like graphs, and most of the rest of the population does not. The majority of college grads seem more adept verbally than quantitatively, although a sizable minority are numerate, and the overwhelming majority of non-college grads are innumerate. So lots of graphs reduce the appeal of a book to the literate but innumerate majority of the reading public.
I'll take a graph over text any day. The "graphs reduce sales" idea sounds like a correlation-equals-causation error. Books with graphs are more rigorous and quantitative; THAT reduces sales.
Graphs are great. On the other hand, equations showing integrals, derivatives and roots make my eyes glaze over unless there is someone to lead me through them very slowly
I love graphs and tables, and I'd guess that so does a large portion of your audience. In your line of work one can't help but attract like-minded readers. You're not writing for magazines that lure buyers with pictures of Tom Cruise on them. If people are following your blog for years on end, it's probably because they share a sizeable portion of your interests.
When leafing through books of quantitative nature, I look throgh the graphs and tables first. If those are interesting, I read through some of the text, if that's interesting, I buy the book.
Oh, and I'm well aware of that Napoleon-in-Russia graph, as are, I'm sure, hundreds if not thousands of your regualar readers. It's a cool graph.
I like graphs, but presentation is important. For instance, I hate pie charts.
Love graphs.
I like graphs (and formulas and statistics). However, most people seem to assume without even trying that they won't understand anything quantitative.
I LOVE GRAPHS
Newspapers and The Economist use lots of graphs. I only hate journo-graphs because they NEVER start the Y-axis at zero.
Clear, concise graphs with easy-to-read legends are helpful to me.
Ive seen some very confusing graphs in my time however, that seemed even muddy the water somewhat. They have to be done clearly to really be effective.
I like graphs
"...did people used to like graphs..."
That's a barbarism. It should be "did they use to like"
I'd sure like a link to the test score graph, I can't find it anywhere on vdare.
Im not sure I understand this. Graphs make information easier to understand. I would think most people would rather just look at a graph than read a whole bunch of words.
A graph can be worth a thousand words.
I like graphs. Walls of texts I start scanning over.
I like graphs.
Meanwhile, I am surprised that so many people are against Power Point.
I like the joke about how a military person defines ellipse (figure from geometry):
ELLIPSE is a CIRCLE inscribed into a [SQUARE 3 BY 4].
Pedantic person’s refutation: ellipse generally is _not_ a circle. There is _no_ such an object in geometry, [square 3 by 4].
However, in Power Point:
You pick a rectangle from drawing menu. By default it appears as a square on the page.
Next you pick an oval from drawing menu. By default it appears as a circle on the page.
Moreover, by the same default, the diameter of this circle is equal to the side length of the above square. As a result, by easy clicks of cursors you perfectly “inscribe” circle into the square.
After that you perform an operation “Group” from drawing menu upon this combination of circle and square.
After that you stretch the resulting group by factor 3 in vertical direction, and by factor 4 in horizontal one.
Just in case, you rotate the resultant object at an arbitrary angle.
After that, you perform the operation “Ungroup” from drawing menu, and then remove the remnants of former square.
Here is your perfect ellipse, in this case with aspect ratio 3 by 4.
I also like graphs, and urge you to keep making them.
Graphs are great, but there are powerful disincentives.
I mean, just consider it from the technical writing standpoint. If I'm writing an essay which is all text, I'm basically done when I've written it. It's immortal; will never bitrot. I can convert it easily to any format I please. It's trivial to store or transmit or... etc. etc.
Suppose I add just one measly small graph. Suddenly I now have *2* files, which makes transmission and storage much more difficult; I need to worry about filepaths so whatever address I use in the essay maps correctly onto the actual filename of the graph image. I have to worry about OSs. And so on and so forth. (It's even worse if I'm using something like LaTeX which I can programmatically specify the graph inside the text itself, because now my text file is source code, and I have issues with generating it, API drift, installing the compiler...)
Backing up your anecdotal evidence about book sales--
I have rejected a couple of books solely because their graphs were shoddy or trying to bamboozle me. It was a really quick way of seeing that the author could not be trusted.
Google Books Ngram Viewer.
Trivia - Raghuram Rajan = South Indian Brahmin. This is probably why he's willing to address issues that others won't touch with a barge pole.
I like graphs and hate equations.
Calculatedriskblog.com is one of the most popular econ sites on the web. It is popular because of its many graphs so some people must like graphs.
First off, graphs have an often outsized influence in society. Florence Nightengale created a polar pie chart demonstrating that vast majority of deaths in the Crimea were from post-surgical infection, and that the deaths in places where hand-washing and clean (boiled) bandages and sheets were used were dramatically lower than where they were not.
Nightengale's graphs were able to convince Parliament to order Doctors to wash their hands and boil bandages and sheets.
David McCandless at TED argues that graphs are the key to avoiding information overload. He does caveat that they are hard to produce. He gives the example of the famous facebook breakup chart, which took him about six months to produce by refining scripts (likely PERL) to crawl facebook pages and record relationship status and calendar day. He found breakups peaked at before Christmas, before Summer, Spring Break, but were lowest on Christmas Day (too depressing and cruel).
McCandless further argues (and I agree) that there is tremendous amounts of data that can ONLY be understood through visualization/graphs/charts, particularly connections and relationships. And that also, Facebook is a tremendous repository of data, mostly untapped, for various social/market research.
IMHO the market valuation of Facebook is probably its ability to tap internally all this data. Far easier than outside crawling web scripts.
FWIW, "Burn Notice" had a nice piece of visual writing that conveyed in an image a key plot point.
A bad guy wanted the flight information for all private planes in and out of Florida for several months. The hero's friend in a nifty piece of exposition, overlays pictures (on a light table) of each day's flight plans, one day only has a giant hole in the flight paths where no other plane comes close.
In the image, the idea is easily grasped, for a broad entertainment audience. It is also dramatic, in that it entices the audience to wonder what sinister purpose is behind the authorities re-routing all the flights for that day only (to keep people away from ... what?)
To convey that in words would have taken both the dramatic impact away, and far too much (limited) time. Other examples are Star Wars wireframe graphics of the rebel moon-base nearly clear of the gas giant in the climax of the film. So clearly TV and movies at least occasionally use graphs/charts to convey quickly complex information to a time-stressed audience.
The negativity around charts is that they are hard to produce (it is easier to write) and there are too many "junk charts" by Powerpoint Rangers designed to obfuscate instead of illuminate. Ironically just recently the FT has had a series on accounting/ERP systems that allow executives to drill down on profitability and cost, from simple bar charts to data underlying (reflecting that just a few customers can supply 80% of profits). Information hidden unless you can access it by chart to chart drill down.
Graphs = 3" graphic representations of data that are far superior to the same data presented in an overly wordy 3 column article. Just the facts, visibly presented and easily assimilated. What kind of an idiot would hate a graph?
I blame the Economist.
You know, the magazine that constantly, hilariously, gets bamboozled by some transparent hoax that insults the kind of Americans that the Economist hates?
Well, that same magazine seems to plot EVERYTHING in graph form, even when it adds nothing to the comprehension of the facts. Like, sure, jets now are a lot more expensive than jets used to be - do I need this data plotted? The Economist thinks so.
I love graphs. It's easy to fool people with text (written or audio), and it's even easier to fool them with dense equations packed with tiny subscripts and Greek letters.
If the media would allow just one good presentation of the right graphs (which have been available and valid for 20 years) the whole Global Warming scam would be over in a minute. That's why the media never allows graphs to be used on this subject.
I'm a nerd, my background is technical, and I hate most graphs because they aren't convincing--I can't see your raw data, so I assume you're lying or wrong. Anyone can lie in powerpoint.
Pretty graphs are difficult to generate. There should be far more noise and outliers in any real graph from any real data than a book publisher would ever allow. So someone's "cleaned up" the data. Maybe they did so well, maybe not, but I can't tell.
The number of times I've read graphs in published books with axes that cannot possibly be what they claim is in the dozens.
Right on, Simon in London! You said it first - start the damn y-axis at 0, with a coupla exceptions. Firstly, if the dependent variable is Temperature, and it's not in absolute (Kelvin or Rankine), then zero means nothing. (this goes for any units for which zero is not meaningful)
Secondly, if you have one graph with y starting at zero, but then want to show a close up of a small change, then it is acceptable to have your y-axis start wherever you want on the 2nd graph.
Most graphs in newspapers, especially USA today, are crap. The other thing I don't like is the naming of the graph when it is just (example) "Salary vs. Years on the Job". We know what the axes are by the axis titles "Years on the Job" and "Salary in $/yr". Put a bit more imaginative name on the top, like, "Salary Changes this Decade", so as not to just repeat the names of the axes.
I luv graphs! A graph is a picture, and a picture is worth a thousand words, as someone said above.
It depends on the graph, don't you think?
Many bar graphs are completely useless, I find. The numbers tell me everything I want to know and more -- in fact, just about all I do with the graphs is to read them for the numbers they depict. This is mostly true for pie graphs as well.
To be useful, the graph has to tell you something visually you'd otherwise have difficulty fully grasping. A scatterplot, for example, can be quite useful in showing things raw data can't -- such as the pattern of outliers, etc.
And graphs are most useful when they depict relationships between two or, especially, more variables that simply can't be easily pictured in the minds eye without them. For example, an economist might in the same graph depict the relationship between private debt, public debt, and economic growth from the Great Depression through World War II and beyond. It's hard to see how anything but a graph can communicate the underlying pattern so well.
"A bad guy wanted the flight information for all private planes in and out of Florida for several months. The hero's friend in a nifty piece of exposition, overlays pictures (on a light table) of each day's flight plans, ... "
Whiskey, I understand your point about the pictoral display of the info., but this is why I don't watch those kinds of shows (besides that the bad guys are always white) - too much BS....
You don't need a flight plan to fly, and you don't need to be on anybody's radar (other than primary, which is sporadic). You don't need to talk to anyone on the radio. It's called Visual Flight Rules, and this is all so long as you stay out of certain pieces of airspace, of course.
So, what were those guys thinking on your show ... only thing that'd work would be "hey, we're in a TV show - let's profile the white guys." ;-)
Regarding the sales halving of graphs, Stephen Hawking wrote in the introduction to A Brief History of Time that his publisher told him the same thing about equations.
Graphs require visio-spatial skills, ie. rotate a 3 dimensional object in free space. Visio-spatial is like the engineer-G. People seem mostly born with a certain aptitude and those that have high visio-spatial excel at engineering & math regardless of the type of engineering & math they are actually doing.
But not everyone has it. For people with low visio-spatial skill, looking at a graph is like reading something written above one's "grade level". It goes in one eye and out the other, so to speak.
Mel
The show's in Miami, and the area was 12 miles off shore, so you're gonna need that flight plan (or you'll get some company along on your VFR tour.
And graphs are most useful when they depict relationships between two or, especially, more variables that simply can't be easily pictured in the minds eye without them.
Right, Wynne Godley's Sector Financial Balances equation is hard to visualize mentally.
Domestic Private Sector Financial Balance + Fiscal Balance + Foreign Financial Balance = 0
But show a graph of how the sectors all balance in several different national economies, its much easier to understand.
http://media.ft.com/cms/fd3b7774-789f-11df-a312-00144feabdc0.gif
No, but everyone openly detests the shoddy Vdare website. It's time to stop your own little welfare to that site. Poorly designed, poorly linkled, etc. PB says it's a 24/7 job- bullshit, utter bullshit. Tell him to get it together or let it fall, because right now that sight it utter crap.
"...did people used to like graphs..."
That's a barbarism. It should be "did they use to like" --anon
Correct. "To use" should behave like any other verb in this spot-- you wouldn't say "did people liked graphs?"
But your point would have been clearer had you graphed it...
Clearly large swathes of the great unwashed dont like graphs, but some of us do. This does impact on book desgin and publication. Too many graphs - or too few.
It would be handy to have a convenient way of looking at these relationships, hey...how about a nice graph?!
I stand corrected, Bluto, thanks. I guess, since I don't watch much TV, I have never seen this particular show. I didn't know it was set in Miami (Whiskey just said "coming into Florida"). Of course, if you're up to no good, you definitely don't want to file a flight plan ....
The simple reality is that people absorb information in different ways. Some are audiophiles. Others are text oriented and some are visually oriented. As the saying goes, “a picture is worth a thousand words”.
Rajan, who is probably text oriented, simply left out the graphs because he assumed everybody absorbed information the way he does. I feel the same way about maps as I do graphs. Whenever I read, I always have an atlas handy. It also irks me that many books will pepper the maps throughout the book rather than in one area where they can be easily referenced.
Matt Taibbi hates Tom Friedman's graphs:
http://www.nypress.com/article-19271-flat-n-all-that.html
Graphs, at least the ones you do, are a good thing. Cartoons would be even better. Preferably by a right-wing cartoonist who uses simple, bold lines:)
Worst use of graphs ever.
Gelman of Columbia U.
last page of document
logarithmic scale with UNLABELED axes.
It's like he is intentionally obscuring the data so you don't notice how bogus his work is.
A report only the NYTimes could love.
Looks like this is it for VDare. I understand Brimelow's sentiment to some extent, but that appeal is over the top, and riddled with typos. And I wanted to see the graph.
Kind of uncharacteristic of Mr. Brimelow. Did he hire one of those ubiquitous English major girls to run VDare or something? That's been the bane of many a publication, but a lot of the old timers seem to think it's a good idea for some reason...
VDARE is back up but where is Steve's graph?
Where in VDARE is that graph?
Accurate and elegant graphs are fun.
But it would be more informative (unless one is preaching to the converted) simply to give the words and numbers from which the graph is drawn.
Another objection is that graphs make deception easier. Words refer directly to things, so BS is easier to detect in text (ditto numbers).
Every middling manager groks PowerPoint because graphs perfume and befog BS with an air of: "This is Science." No, it isn't; it's esthetics.
I LOVE your graphs. They make your very complicated arguments and presentations of large amounts of data MUCH easier to understand. Most people make a mess of graphs. YOURS are a joy.
Please keep including them!
Thanks,
David Rogers
http://www.the-american-interest.com/article-bd.cfm?piece=906
The problem is not so much 'plutocracy' as 'Jewtocracy', and it's especially bad for white people.
Among blacks and Hispanics, there is a connect between elites and masses. Most black elites are Democratic and most black masses are Democratic. Both are into black power and interests.
Most brown elites are Democratic and most brown masses are Democratic. Both are into brown power and interests.
But because Jews count as 'whites', there is a disconnect between white elites and white masses. Jews are overwhelmingly liberal-left while majority of whites are center-right. Since Jews constitute the 'white elite', white people have an elite that works against their own interests. Of course, Jewish elite and Jewish masses are well-connected(though it seems nearly 40% of Jews are socio-economically of elite category). Jewish elites serve the interests of Jewish people. But because of the dual nature of Jewishness--as both white and non-white--, Jews are seen as the 'white elite'.
In addition, while Jewish elite admonishes white people of 'white guilt', Jews present themselves as the primary victims of white evil--especially the Holocaust. So, we have a strange set of affairs where the rich & powerful white elite(Jews) are seen as the victims of underprivileged white masses. What can be more screwy than that?
It's not that everyone hates graphs, it's that you make hategraphs.
Post a Comment