Data Story: Napoleon vs. the Alpha Male

Social scientists at Texas Tech university have recently published their findings about voter preferences for taller leaders. To test this hypothesis, last week the Guardian put together a list of 51 international statesmen and women, and their respective measurements. These findings invite all kinds of speculation, so before drawing any conclusions, let’s look at the distribution of the sample, weed out a few anomalies and apply some subjective filtering.

The Guardian’s data sample ranges across eras and cultures, from the diminutive Mexican president Benito Juárez (1806–1872) to Gallic man-mountain General Charles de Gaulle (1890–1970), with many historical and contemporary figures in between. The sample size is relatively small, and one feels they’ve deliberately chosen some outliers, but the selection is spread over a wide range of countries and eras.

Twin Peaks

In a normal population, heights would be distributed into a bell curve. By contrast, this group is clustered into two peaks, and there is a big gap around about where you’d expect the apex. A slight majority of leaders fall below the mean height of 5’9″, with a long tail at the lower end. Taller leaders peak in the top quartile, at 6’1″.

Gender correction

While it’s possible that these rules also apply to women in office, there’s no getting away from the fact that the female of the species are on average smaller in stature, so including them in the sample is going to skew it downwards. Hence, we can remove five ladies of decidedly average height, and one a little above it. So while men of average height are not at all represented in the sample, neither are there any women significantly outside the norm.

Leadership Style

Let’s remove some of those troublesome dictatorial types, who achieve power by coercion or other non-democratic means. A total of ten men, only three of them above average height. This is a qualitative discrimination, by which the definition of democracy gets a stretching and we remove candidates who may be guilty of perverting the democratic process. For instance, Iranian premiere Mahmoud Ahmadinejad came to power in 2005, and was re-elected in 2009, but the latter of these two victories has been called into question. Likewise, Vladimir Putin has been elected by the people in the past, enjoys considerable popular support, but has lately displayed a certain disregard for the ballot box, and currently holds a position of his own devising. And we all know who else was elected to power in 1932.

Public Preference

But wait. By the same reasoning, we can arguably filter out some other people who didn’t actually win elections. George Bush fiddled it at least once; David Cameron and Nick Clegg both failed to win outright; and Gordon Brown and James Callaghan both came to power without having been elected by the public at all. All men of above-average height, who don’t exactly qualify as dictators, but weren’t necessarily preferred by a voting majority either.

At this point, we may have homed in on voter preferences, but we’ve lost more than a third of the sample, and the data no longer looks like very much of anything. The overview suggests that while the public may indeed prefer to put an alpha male in office, those of smaller stature are just as likely to put themselves there. But the two smallest men in the sample were respected statesmen with impeccable democratic records. Conspicuous by their absence are leaders in the middle height range, but norms haven’t been corrected for geography, time period, economic conditions or living standards.

Tempting as it is to buy into our expectations about the nature of power and the physical stature of the ruling elite, apply a few specious arguments to a tiny sample selected with an editorial eye, and the information can mean anything at all. With this in mind, we submit the following absurd but strangely plausible hypotheses:

  1. Men of average height are underrepresented in high political office.
  2. So too are women of any height.
  3. A dictatorial style of leadership is often found among shorter men.
  4. Taller men have been known to game the system.
  5. Great statesmen can be very short indeed.

But perhaps the most striking conclusion is this: if voters do in fact prefer taller leaders, this is not evidenced by any process of selection as demonstrated here. Depending on what you believe, this might be another way of saying that if democracy exists as a system for expressing the wishes of the electorate, then it doesn’t actually work.

 

Dynamic data visualization. Why bother?

We are creating and saving more information now than at any point in human history. At work and at play, in ways that nobody controls, we produce and store massive swathes of data and metadata. These resources often contain fascinating stories and unexpected patterns, surprises which inevitably elude us if nobody investigates. But before we decide what to look for and where to go looking for it, shouldn’t we be asking: why bother?

Navigating the ever-rising tide of information is something of a concern, so you could be forgiven for thinking that everything from your extended social network to your laundry list should be subject to analysis by default. But for every paean to elegant visual precision, there can be untold hours of unglamorous data entry, cleaning and wrangling. Add to that the skill required to interpret and model the outcome, and you begin to appreciate that even with the right tools at your disposal, conclusive visual analysis doesn’t come cheap, particularly if you don’t know what you’re looking for.

Read More »

Interview for Digital Arts magazine

REPOST: Earlier this year, we were interviewed for Digital Arts magazine, about the Budget Explorer application we’ve been building for Twaweza in Tanzania. The platform serves as a one-stop shop for exploring public spending.

How did the project come about? What were your aims?

We built a prototype visualization for the British project Where Does My Money Go? which was based on UK spending data published by the Treasury. Rose Aiko, research analyst at Twaweza saw this, and contacted us with the intention of developing something similar with the Tanzanian budgetary data.

The idea is very simple: to explain how government spends its money, on a national and regional level. This can have  tremendous value not only in public discourse, but also to policy makers and activists, because it makes it much easier to track how monies are being allocated, and whether budgeted resources are are being targeted effectively.

Read More »

Plings users need book-marking tools

REPOST: This post originally appeared on the Plings.net blog

One of the hardest things about designing rich applications for the Internet is discovering what end users really want, and how to deliver it. It’s essential to have some idea of what your customers are looking for if you want to stay in business, but how can you provide an experience which really clicks with your target audience, and makes them come back again and again? Delivering effective results means communicating with an audience about what they want from the highest level, right down to the smallest detail.

In designing Plingorama, we’ve been talking to young people about what they want, in a small focus group and watching people use the application as part of usability testing. This is what we found out.

Read More »

Follow

Get every new post delivered to your Inbox.