For me, it follows in the mould of The State of Obesity, but is much more transparent in how it is constructed when you look at the source code. That makes it a good exemplar — in fact, perhaps the exemplar currently available — for introducing people to the possibility of making interactive dataviz for their research projects.
Oh for those early days of D3, when nobody was terribly fluent with it, and websites would have all the code right there, easy to read and learn from.
That transparency is important, not just for teaching about dataviz, but for the whole community making and innovating interactive data visualisation. Oh for those early days of D3, when nobody was terribly fluent with it, and websites would have all the code right there, easy to read and learn from. Now they are tucked away in obscure links upon links, uglified and mashed up with other JS libraries, familiar to the maker but (probably) not you. There are obvious commercial pressures to tucking the code away somewhere, and you can actually obtain software to obfuscate it deliberately. At the same time, having everything in one file, hard-coded for the task at hand, may be easy to learn from, but it isn’t good practice in any kind of coding culture, so if you want to be respected by your peers and land that next big job, you’d better tuck it all away in reusable super-flexible in-house libraries. And yet, the very availability of simple D3 code was what kick-started the current dataviz boom. Everyone could learn from everyone else really quickly because everything was open source. I don’t like to think that was a short-lived phase in the early part of the technology’s life cycle, but maybe it was…
Anyway, that’s enough wistful nostalgia (I learnt yesterday that I am the median age for British people, so I am trying not to sound like an old duffer). Here’s the things I don’t like about it:
- it requires a very wide screen; there’s no responsiveness (remember your audience may not work in a web design studio with a screen as big as their beard)
- life expectancy gets smoothed while the other variables don’t – just looks a bit odd
- why have colors for continents? Doesn’t it distract from the shading? Don’t we already know which one is which?
- Why give up on South Sudan, Somalia (which seems to be bunged together with Somaliland and Puntland in one big “hey it’s somewhere far away, they won’t notice” sort of way) and North Korea? Aren’t these countries’ estimates kind of important in the context, even if they are not very good? Do you really believe the Chinese estimates more than these just because they’re official?
But all in all, these are minor points and a nice amount of grit for the mill of students thinking about it. I commend it to you.