10 ms·
Mistakes, we’ve drawn a few
- nerdponx 8y agoOn the dog weight and trade deficit charts, I'd argue there's merit to the originals. The essence of a linear relationship between two data series X and Y is that X and Y are identical up to position and scale. The "better" charts represent the data more accurately, but the original charts make it easier to see the relationship between the two series. Also the trade deficit chart is confusing in part because the negative sign isn't obviously a negative sign. Edit: the last chart is fine. It's maybe not very sexy, and it takes a moment to absorb, but the data is clearly displayed.
- kbenson 8y agoThe dog weight one is a good case of how you can futz with numbers to imply a greater relationship than is present. I think the second chart is much better in that respect. If you're doubling the relative range of one of a double scale to make stuff look like it fits, that's not exactly being forthright. > but the original charts make it easier to see the relationship between the two series. Well, as the article says, if it looks like they're a perfect fit when they definitely aren't, that's probably beyond making something clearer and going into convincing people of something that the chart doesn't really imply. > the last chart is fine. It's maybe not very sexy, and it takes a moment to absorb, but the data is clearly displayed. Totally agree on this. It takes a bit to digest, but that's because there's a lot of information imparted.
- function_seven 8y agoThere's no reason the scale needs to be the same when comparing measurements in totally different dimensions. The weight is roughly a volume measurement while the neck size is a linear one. Also, if the necks were measured in inches, you'd end up at roughly the original chart, even being mindful of the scaling on each axis. I think the left chart is actually the better one. It does a good job showing the correlation between neck size and weight.
- Quenty 8y agoIt’s a matter of percent change. It shows the same relative percent change on the scale, which feels more accurate. I prefer the revised graph.
- gowld 8y ago> > The weight is roughly a volume measurement while the neck size is a linear one. so "percent change" is a different dimension in the two data sets, respectively.
- ovi256 8y ago>The dog weight one is a good case of how you can futz with numbers to imply a greater relationship What they did was very similar to standardizing data series (normalizing to mean=0 and standard deviation=1). This is a standard tool in scientific data analysis (most published papers use it) and it's taught in statistic and data science classes. It's also a preliminary step in computing most correlation measures. It does not exagerate relationships, it helps visualizing them. The second chart removed the standardization and obfuscated the relationship.
- SiempreViernes 8y agoAs you can clearly see from the axis labels, they are not plotting standardised or indexed data, what they did was hand picked a rescaling that makes the plots line up very nicely. Further, it is simply not true that weight is a perfect function of neck size, so you shouldn't make a plot that implies that.
- gowld 8y agoThe reason it's hard to pick which chart is better is that both charts are far too detailed for the tiny amount of information they convey, which is "both values shrank over time"
- SiempreViernes 8y agoCan you further explain the frankly astonishing statement that you find it hard to pick between a plot showing a false equivalence and one that doesn't?
- ovi256 8y ago>what they did was hand picked a rescaling that makes the plots line up very nicely I bet dollars to donuts that's what a stardization would show too, unless I'm missing something. It would show an even more perfect lining up because it would line up the line ends. Wish I could test it. Wish pandas had a read_image function to get data from line plots like it has read_html.
- soVeryTired 8y agoAs you suggest, it's not always incorrect to shift and rescale data for visualisation. In fact, that is exactly what a linear correlation calculation does: the data is shifted to have mean zero and rescaled to have variance one, then the covariance is computed. The real problem is that often, the two timeseries being compared are too simple to demonstrate whether there is a relationship present. If they're both essentially straight lines as in the Butcher's Dog graphs, you may see a spurious relationship. What you need -- in unscientific terms -- is to see one timeseries do something weird, and the other timeseries reacting to the weird thing in a manner that fits the narrative. Straight lines don't really do that, but often the first difference of two timeseries that look like a straight line will be good enough.
- ukoki 8y agoI don't get the dog one at all. The article is about dogs getting smaller. Since neck size and weight are very correlated why bother to show both? I would just have a single plot showing the decline in dog weight over time. The fact that smaller dogs also have smaller necks is neither surprising nor useful.
- spectramax 8y agoI applaud the fact that they admit to the mistakes and owning up to them. Converting abstract relationships into semiological representation is not exactly well cut science. There is a bit of experience and discovery in being able to take complex data and present it in a clear manner. Now, only if we can get publications like Bloomberg to admit their mistake about "The Big Hack" but off-topic and I digress. I've always enjoyed Economist and its British humor, and this article just adds more credibility.
- hi5eyes 8y agodont most people pass off the bloomberg hack story as stock manipulation
- mikekchar 8y agoPersonally, my tinfoil hat theory is that it was concocted by the white house staff to create a political climate conducive to playing hardball the Chinese. Not one shred of evidence, of course, but it seems the most plausible thing I could think of.
- chiaro 8y agoOr the classic least norwegian squares method of regression in the WSJ http://kieranhealy.org/files/misc/laffer.png http://kieranhealy.org/files/misc/laffer.png
- lwansbrough 8y agoI really like the third graph redesign -- the one with the average/trend overlaid on top of the fainter data points. Is there a name for that?
- haberman 8y agoThis is a confusing graph from the Economist that I snipped when I saw it in 2013. Very difficult to wrap your head around what is being shown here: https://pasteboard.co/I7ttkG5.png https://pasteboard.co/I7ttkG5.png
- jacobush 8y agomy browser complains about SSL for that link
- isolli 8y agoFirefox tells me "Image not found", but Chrome displays it happily...
- majewsky 8y agoWorks fine for me in Firefox.
- schoen 8y agoIf you can show me the specific error, I'd be interested to look into this, since I work on Let's Encrypt, which provided the certificate for that site (and it's useful to know about browser compatibility problems or new potential kinds of misconfiguration).
- jacobush 8y agoSecure Connection Failed An error occurred during a connection to pasteboard.co. Peer received a valid certificate, but access was denied. Error code: SSL_ERROR_ACCESS_DENIED_ALERT The page you are trying to view cannot be shown because the authenticity of the received data could not be verified. Please contact the website owners to inform them of this problem.
- B1FF_PSUVM 8y agoIt's not a bad graph, it is just not familiar. Perhaps it would be more clear sticking to a single value of US GDP, e.g. as "% of world pop. above 1/4 of US GDP per person, over time", but the data shown only has three time points ... (Thanks for the pic, it was interesting.)
- kurtisc 8y agoI've noticed the scales issue happens a lot with claims about markets. A story will say something like 'Asset A plummets among bad event X' with a daily graph of the price. When you look back at the weekly or monthly graph, though, you find that it's well within normal fluctuations and isn't even a low for that period.
- openloop 8y agoA few? What is more concerning that chart errors are clearly conservative viewpoints with completely misleading data. Half of the time this paper is unfit to print.
- dmd 8y agoIt is a family tradition in my house to pore through the Economist every week looking for the worst visualization. So, props to them for admitting they have a problem.
- Waterluvian 8y agoI do the same with CBC. The offense I hate the most is when they chop off a bar chart so one bar is twice the height of another but the actual value is only like 5% higher.
- dx034 8y agoWouldn't you find more in other newspapers? I generally find Economist visualizations surprisingly well made. Articles like this show that they're actively trying to improve.
- gowld 8y agoIf it's not a challenge, it's not fun.
- rwmj 8y agoI'm an Economist subscriber and so read it most weeks, but honestly their charts and visualisations are often terrible. I think they could read a bit of Tufte and reduce chart junk as a starting point. This article does make some other good improvements though, so there is hope.
- samfriedman 8y agoFor more reading on poor viz & better redesigns, I recommend Junk Charts by Kaiser Fung: https://junkcharts.typepad.com/ https://junkcharts.typepad.com/
- itamarst 8y agoMy favorite mistake The Economist made was a survey they ran. They compared UK to US, and supposedly one of the questions was: What better explains the origins of the Earth? A) Evolution. B) The Bible. It's unclear whether they literally asked that question, or they asked about origins of life and just had a copyediting mistake in the summary of the results. They ignored my letter about it, though, so I suspect the former :)
- mfoy_ 8y agoThat's funny, because technically the answer is B.
- SketchySeaBeast 8y agoIf does provide a cleaner answer, but only in the same way that the flu can be more easily described as a witches curse.
- ysavir 8y agoTheir point is that evolution doesn't explain the origin of Earth, a planet, at all. While the Bible's explanation may be fiction, it does offer an explanation for the origin of Earth, and is therefor the better explanation by virtue of providing an explanation.
- mfoy_ 8y agoYeah, with the way it's phrased, it's a trick question.
- SketchySeaBeast 8y agoAh, absolutely true. Evolution doesn't explain the origin of the cosmos. My lazy brain fell right into the trap.
- mcguire 8y ago
- oritron 8y agoThe most surprising conclusion was about the final plot discussed, regarding scientific publications. The author says that it contains too much information and is confusing for that reason, and postulates it might be better to summarize the different fields and plot as a single variable. The truth is, it is not a confusing visualization. There is no rule that says to remove variables for the sake of reducing a variable count. It is useful to see that health sciences stands apart from other fields, and other points can be seen as a cluster or examined more deeply by the reader. It is actually one of the best plots shown in the entire article. In effect it is a grouped bar chart, a visualization that many publications shy away from in favor of stacked bars. While that is good for showing a total, it is often a less effective way of visualizing data—it's difficult to compare magnitude of anything but the base bar. That fault can be seen in the stacked bars of the budget surplus example, which still exists for the improved version. At any rate, I'm glad the Economist is looking to improve their visualizations and isn't afraid to share mistakes.
- SiempreViernes 8y agoThe precise statement was that at that size it was confusing and hard to read, which I think is a fair statement.
- mcguire 8y agoMy impression is that there were too many countries shown. (Possibly they were mentioned in the article, and therefore relevant to the context?)
- auraham 8y agoI enjoyed the post. I am not familiar with the website, but from the examples in the post, it seems that only the x-axis is labeled. It can be hard to understand the meaning of a plot when only one axis is labeled.
- minikites 8y agoMy college statistics textbook had an entire problem set where every question was dedicated to a bad or misleading chart from "USA Today". These are mild in comparison, but it's important to get visualizations correct, given how powerful they are at communicating ideas and concepts.
- cartercole 8y agothey did ben so dirty :(
- arc_of_descent 8y agoI'm currently reading Statistics for Dummies, and these kind of techniques used i.e. wrong scale, incorrect starting point, etc is explained very well. Highly recommended.
- pcvarmint 8y agoSuch as calling Ben Shapiro an "alt-right sage"? [0] 0. https://www.economist.com/open-future/2019/03/28/inside-the-mind-of-ben-shapiro-a-radical-conservative https://www.economist.com/open-future/2019/03/28/inside-the-...