While I appreciate the perspective on the relative investment in science, SciAm, this is a terrible graphic.
It’s interesting that the total American investment in science is slightly larger than what the military throws away on just the F-35 program, but otherwise, this image has very low information density.
Humans are lousy at comparing surface areas, and here the money spent on each project is represented by…the area of a circle. That’s poor data representation. I’m sorry, but the creator of this graphic is going to be lined up against the wall with the person who invented pie charts.
Each data point is shown as a circle drawn arbitrarily somewhere on the page. There is no information in relative location — I guess the circles were just splotched down in an arrangement that looked good to a graphic designer somewhere. Are there any relationships between any of these data?
Colors are also arbitrary. Nations are in blue, genomes are orange, brains are purple, and telescopes are red. Just because.
There is so little information in this wall of circles that it needs to be helped along with paragraphs of text dumped onto the page in teeny-tiny print, and most of that text doesn’t tell us anything about the relationships between the circles. They are just circles of varying size on a bland blue background.
A tidy table would have been more than adequate for expressing numbers.
Oh, and there is content. Let’s shut down the F-35, give all that money to the NSF, and double the amount of grant money.
Worse yet, they compare funding for different numbers of years in different circles. Maybe they should have used apples and oranges instead of circles.
I saw that a bit ago, took a look, and gave up.
Glad someone didn’t just give up.
The circles for countries are arranged linearly from largest (USA) to smallest (Saudi Arabia. Presumably, those are the largest recent national expenditures for scientific research, but there is no representation for the total of all other countries combined. Apparently, the costs of developing various technologies is also included, or other things that may fall into that general category in the official budgets of different countries, as if that counts as “science” in a strict sense. Who knows, perhaps it also counts spending on things like alt med or pseudoscience.
They give you such nice, helpful circles for alcohol, Avatar and the F-35. That makes some kind of a statement, but it is not even an attempt at giving you anything like a general perspective on “how highly humanity as a whole actually regard[s] science,” compared to all sorts of other things, just a few dollar figures with no coherent sense of context. Even if we took dollars as our measuring stick, how big is public funding for the arts, for instance? Or what about funding for all non-scientific academic research? How many pixels would that be? Never mind … apparently, it doesn’t matter.
Looksh purrrfectly cleer too moi! Wobbles off blearily.
As far as I’m aware the Pie Chart was invented by an eighteenth century Scottish economist called William Playfair, who was in Paris at the outbreak of the French Revolution and took part in the storming of the Bastille. So, in a sense, he was up against the wall entirely by his own volition.
Are the circles scaled simply by gross expenditure by nation? Could they scale it by each nations GDP? or minimally, per capita? How about Percentage of National Budget?
They could have done that. Would that be better? They offer this note:
They’re not giving you any relative information, comparing these numbers in a useful way to some other set of numbers. How much money is there in the world? How big is that circle? Nowhere to be found. They are absolute figures of simply how much is spent, period, albeit ones that are adjusted to factor out (certain kinds of) irrelevant economic differences.
That also doesn’t help with #1’s problem, that a single-year budget is compared side-by-side with specific programs that last over the course of 10+ years (or more? some are projections about programs that aren’t complete, haven’t even started doing research, might be delayed or extended for much longer, etc.). They’re also generally funded by one or more nations, so they should be inside those nation’s circles (or spread out into several if that’s the case, to make it maximally confusing), except that they’re citing some old things like the Apollo program that have nothing to do with the USA’s 2012 budget. Or is that not only public spending in the US? Is it private spending too? Does that not count? What about the case of that country over there, which does things differently?
I’m going to nitpick. A pie chart isn’t meant to represent area but a *linear* proportion of a circular whole in relation to the *thickness* of a slice It is a purely linear thought process and humans are quite competent at it and, so far as I can figure, is the best way to visualize relative sizes of fractions (linearly).
But I agree that humans are terrible out comparing areas. They are good at linear comparisons. But they are then bad again at seeing a bunch of linear data and comparing each bit to a total whole. Which is why pie-charts are *good*. The compare data *linearly* but also while giving a visual framework of the whole.
Alternative lesson from that graph for Republicans: Look at how much money we’re wasting on “science”. Let’s take that all away and double the budget of the F-35 project.
The F-35 is already dead; it’s just that the lights haven’t been turned out. The aircraft is notably inferior to all the aircraft likely to fly against it, which means it’s flying lunchmeat. Furthermore, the CAD files for the thing ‘leaked’ out* and so any nation that wants to build a radar cross-section test dummy can make sure its alleged ‘stealth’ can be defeated. The only think keeping it alive is that it’s a gigantic jobs program.
The littoral fighting ship and the Osprey are two other gold-plated boondoggles that need to be killed- and fast. Do you see a pattern here? Each branch of the military has its own gold-plated commode that they flush taxpayers’ dollars down. And of course the intelligence service has theirs, too…
If we aligned our military spending to something reasonable, we’d be able to do so much useful good work with the money, except it’s better to funnel that into the hands of rapacious executives in the defense/industrial complex (and there’s a minor “trickle down” effect by creating jobs in the dark tech sector)
The brits have these rituals involving fucking dead pigs, to get people to buy into the oligarchy. In the US, the test is to buy someone and see if they stay bought. There are a huge number of people who watch the defense/industrial complex keep ballooning out of shape and remain silent because their $100k/yr job is pretty good and, well, they’re getting theirs…
You’re absolutely right about surface area being a bad way to present data. When I teach classes on metrics, one point I try to make is that pie charts and blob charts are deceptive because they present ratios in effect at a given slice of time. If you look at a time-series chart you are looking at a load of pie charts, where each pie chart is a vertical slice through that chart at the instant of the slice. When someone shows you a pie chart, they are choosing the moment in time for you, and their motives should always be suspect. Why did they go through the trouble to collapse one axis around that particular time? A time-series showing what programs were absorbing what amount of resources over time would be much much more interesting especially since you could see what was growing or what was shrinking.
(* it’s a program that has literally bazillions of US and non-US contractors including every defense contractor in NATO working on it. for all that there has been a concerted attempt to blame China for stealing the data, it would have been easier to keep the data secret if they had just hosted it on an anonymous FTP site, somewhere)
The F-35 is a turkey, but it will never get defunded even if the military brass give up on it, which they almost have, because the F-35 has parts made in multiple congressional districts.
And yes, all the F-35 money would be better spent putting white plastic roofs on all buildings in the warmer states and building anaerobic cogeneration gas plants on all small town sewer systems. Plenty of money left over for tax rebates for rich people to build personal solar which will aid in the development of better industrial solar.
It totally sucks that Germany, a solar unfriendly country, generates more watts of solar per capita than the US. Is Merkel showing that Hitler was right, that the Germans really are a superior race?
the Pie Chart was invented by an eighteenth century Scottish economist called William Playfair
I immediately wondered if that was the same busy boy who invented the Playfair cipher. It wasn’t.
To those of us who are comfortable thinking in pure numbers, as I guess PZ must be, I’m sure a table would convey as much info and look tidier. To those of us who enjoy a more graphical mode of thought, a graphical representation works better and it’s possible for us to walk away with as much info as the table-watchers.
SciAm is a pop-science magazine, not a respectable journal, so a mere table, presumably with a minimum of “arbitrary” colours and other graphical devices, is just not going to cut it in the pop-sci market. It has to look attractive to the fickle news-stand browsers, and looking like a year-end report for a particularly boring accountancy firm isn’t going to do the job.
Finally, I have to defend my fellow graphic designers here, and point out that most of us know little about science and can only make other people’s data look pretty. If the data sucks that’s down to the journalist who supplied the data, not the pixel-stained technopeasant who was trying to make the badly-researched jumble of data look presentable. Blame that guy, not the artist.
Oops, just read the small print: make that “Blame that person”. One day I’ll manage to join the rest of you in the 21st century, I’m sure.
It’s a caterpillar, right?
It could be a very hungry caterpillar.
I agree that this is terrible, but I quite like pie charts. They can show you some things pretty impressively, like when one value takes up most of the pie. For giving percentages they are really handy and allow you to parse some basic information quickly.
It’s also dangerously close to creating an Ebbinghaus effect in the viewer, especially in the apparent size of the ISS circle. Also I didn’t even see the F-35 and alcohol circles when I first looked at it, because the contrast with the background is so low. There’s so little rhyme or reason to the layout of the circle, I assumed the slightly different colors in the lower right were also merely decorative.
I picked up my current issue of SciAm last night and turned to an article titled “Trouble in Mind.” It’s part of series of articles in this issue on the “State of the World’s Science 2015.” This graph is in the middle of the article, so you would assume it’s related.
The summary under the article’s title starts, “Two years in, a $1-billion-plus effort to simulate the human brain is in disarray. Was it poor management, or is something fundamentally wrong with Big Science?” An “In Brief” blurb (a SciAm feature so people don’t have to read) states, “Yet plenty of blame for the HBP’s woes rests with the project’s funders in Brussels, who put politics ahead of science and exercise poor oversight.”
However, it’s not entirely clear how the graph relates to the article, other than showing examples of Big Science. Usually SciAm articles refer to graphics and illustrations accompanying the article, but this article does not. It’s as if some editor decided they needed an illustration in that spot, and so there it is.
I think this is a fine graphic. The relative sizes of expenditures are easy to compare, and if you want more detailed information the actual numbers are there. Suggesting that the creator of the graphic deserves to be “lined up against the wall” (and then shot, I’m assuming you are implying) for not meeting your standards of information density is an abhorrent.