Posts filed under Graphics (277)

November 20, 2014

Round numbers

Nature doesn’t care about round numbers in base 10, but people do.  From @rcweir, via Amy Hogan, this is Twitter data of the number of people followed and following (truncated at 1000 to be readable). The number of people you follow is under your control, and there are clear peaks at multiples of 100 (and perhaps at multiples of 10 below 100). The number following you isn’t under your control, and there aren’t any similar patterns.

twit

 

For a medical example, here are self-reported weights from the US National Health Interview Survey

nhis-wt

The same thing happens with measured variables that are subject to operator error: blood pressure, for example, shows fairly strong digit preference unless a lot of care is taken in the measurement.

November 14, 2014

Motion and context in graphics

Via Michael Toth  I found this animated GIF from isomorphismes, showing the ‘yield curve‘ for Federal Reserve bonds

tumblr_na17r44bUx1qc38e9o1_400

Michael modified the curve to make it prettier — alternatively, more similar to the style of The Economist.  In both cases, though, I felt the time context was missing.  Using animation rather than multiple plots lets you get a lot more on a page, but you can’t see what’s happening as clearly.

One possibility is to make a separate graphic that shows where you are in time; another is to keep some history by letting the graph leave shadows. In the graph below (based on both the linked examples), there are 12 months worth of shadow lines trailing the solid line, and a grey indicator bar showing where we are in history, with GDP growth and unemployment as context.

yield curve evolution

Even better (though not embeddable in WordPress) would be to make the time axis able to both autoplay and be controllable by the user, as in this example from the R animint package.

 

(update: the code)

November 12, 2014

Africa? Can you be more precise?

From the Telegraph (via many people on Twitter)

ebola

 

Seeing this at the same time as hearing about Bob Geldof’s Band-Aid reboot really emphasises the point that Africa isn’t a single place. The first Band-Aid recording was intended to help people in Ethiopia; the new one is for the Ebola-stricken regions of West Africa. The distance from Freetown to Addis Ababa is about the same as Auckland to Dili in East Timor, or Los Angeles to Bogota (or Addis Ababa to Prague).

On the other hand, the graph does make an important point. Syphilis, starvation, and TB are all very inexpensively treatable. Malaria and HIV are largely preventable, also at low cost. An effective treatment for Ebola will help, especially for medical personnel who are otherwise at very high risk, but in the long run it isn’t going to be enough. If we can’t deliver penicillin effectively, we won’t be able to deliver Ebola drugs. To make a real difference, we need a vaccine that’s good enough to prevent outbreaks.

November 7, 2014

Graphics: automate, then individualise

From James Cheshire, a lecturer in geography in London

The majority of graphics we produced for London: The Information Capital required R code in some shape or form. This was used to do anything from simplifying millions of GPS tracks, to creating bubble charts or simply drawing a load of straight lines. We had to produce a graphic every three days to hit the publication deadline so without the efficiencies of copying and pasting old R code, or the flexibility to do almost any kind of plot, the book would not have been possible.  So for those of you out there interested in the process of creating great graphics with R, here are 5 graphics shown from the moment they came out of R to the moment they were printed.

That is, good graphics rely on both soulless automation and creative design flair. Graphic designers shouldn’t need to put the data in by hand; they should be starting with the output of well-designed software and working from there.

November 6, 2014

State lines

Two very geographical graphics:

From the New York Times (via Alberto Cairo), a map of percentage increases in number of people with health insurance in the US.

insured-map

This is a good example of something that needs to be a map, to demonstrate two facts about the impact of Obamacare. First, state policies matter. That’s most dramatic in this region from the right-hand side, about halfway up:

insured-highlight

Kentucky and West Virginia implemented an expansion in Medicaid, the low-income insurance program, and had a big increase in number of people insured. Neighbouring counties in Tennessee and Virginia, which did not implement the Medicaid expansion, had much smaller increases.  The beige rectangle at the top left is Massachusetts, which already had a universal health care law and so didn’t change much. (Ahem. Geography and orientation apparently not my strong points. Massachusetts didn’t change, but that’s Pennsylvania, which only just started Medicaid expansion)

Second, there was a lot of room for improvement in some places — most dramatically, south Texas. The proportion of people with health insurance increased by 10-15 percentage points, but it’s still below 40%.

 

As a contrast, the Washington Post gives us this,

venn

which is, hands-down, the least readable marriage equality map I’ve ever seen.

 

November 5, 2014

US election graphics

Facebook has a live map of who has mentioned on Facebook that they had voted (via Jason Sundram)

facebook-voted

USA Today showed a video including a Twitter live map

twitter-elections

These both have the usual problem with maps of how many people do something: there are more people in some places than others. As usual, XKCD puts it well:

xkcd-elections

Useful statistics is about comparisons, and this comparison basically shows that more people live in New York than in New Underwood.

As usual, the New York Times has informative graphics, including a live set of projections for the interesting seats.

 

October 18, 2014

When barcharts shouldn’t start at zero

Barcharts should almost always start at zero. Almost always.

Randal Olson has a very popular post on predictors of divorce, based on research by two economists at Emory University. The post has a lot of barcharts like this one

marriage-stability-wedding-expenses

The estimates in the research report are hazard ratios for dissolution of marriage. A hazard ratio of zero means a factor appears completely protective — it’s not a natural reference point. The natural reference point for hazard ratios is 1: no difference between two groups, so that would be a more natural place to put the axis than at zero.

A bar chart is also not good for showing uncertainty. The green bar has no uncertainty, because the others are defined as comparisons to it, but the other bars do. The more usual way to show estimates like these from regression models is with a forest plot:

marriage

The area of each coloured box is proportional to the number of people in that group in the sample, and the line is a 95% confidence interval.  The horizontal scale is logarithmic, so that 0.5 and 2 are the same distance from 1 — otherwise the shape of the graph would depend on which box was taken as the comparison group.

Two more minor notes: first, the hazard ratio measures the relative rate of divorces over time, not the relative probability of divorce, so a hazard ratio of 1.46 doesn’t actually mean 1.46 times more likely to get divorced. Second, the category of people with total wedding expenses over $20,000 was only 11% of the sample — the sample is differently non-representative than the samples that lead to bogus estimates of $30,000 as the average cost of a wedding.

October 13, 2014

Herald data blog starts

The Herald’s Data Editor, Harkanwal Singh,  announces the online site’s new ‘Data Blog’, with the first new post being a map of NZ internet affordability created by Jonathan Brewer.

This has got to be a Good Thing for data literacy in the local media.

October 8, 2014

What are CEOs paid; what should they be paid?

From Harvard Business Review, reporting on recent research

Using data from the International Social Survey Programme (ISSP) from December 2012, in which respondents were asked to both “estimate how much a chairman of a national company (CEO), a cabinet minister in a national government, and an unskilled factory worker actually earn” and how much each person should earn, the researchers calculated the median ratios for the full sample and for 40 countries separately.

The graph:

actualestimated

 

The radial graph exaggerates the differences, but they are already huge. Respondents dramatically underestimated what CEOs are actually paid, and still thought it was too much.  Here’s a barchart of the blue and grey data (the red data seems to only be available in the graph). Ordering by ideal pay ratio (rather than alphabetically) helps with the nearly-invisible blue bars: it’s interesting that Australia has the highest ideal ratio.

ceo

The findings are a contrast to foreign aid budgets, where the desired level of expenditure is less than the estimated level, but more than the actual level.  On the other hand, it’s less clear exactly what the implications are in the CEO case.

 

October 7, 2014

Marriage equality maps

The US Supreme Court declined to review seven same-sex marriage decisions today. The StatsChat-relevant aspect is the flurry of maps this prompted:

I think the New York Times (via Twitter) is my favorite version: the square statebins use geography just as an index to make states easier to find, and (in contrast to the last statebins I linked to) they’ve moved Alaska to the right place

BzS1Q2zIQAAxW3Q

 

(more…)