November 19, 2016
Originally posted on Medium
In the aftermath of the US election, one thing that we have to face is that the relationship between the general public and factual data is pretty tenuous right now. There’s very little trust in institutions and their findings, or in information in general.
This means that our ability to communicate data in a trustworthy manner is increasingly important. The good news: with the internet and interactive media, we have some of the best tools we’ve ever had for communicating complex concepts. The bad news: we’re not really doing it.
Right now, our best tool for communicating fact in online media is probably the infographic. It’s visual, can be interactive, gets a point across, and, most importantly, isn’t a wall of text (like, you know, a blog…). The problem is that, at their best, infographics are both shallow and didactic. They’re about getting a single idea across, and they’re more interested in telling you the author’s opinion than in listening to what you want to know. They care more about stripping away complexity rather than making that complexity manageable. Most importantly, they’re not interested in what questions the viewer has, only in the questions that the author wants to answer.
And when there are so many infographics out there, many of which are definitely inaccurate and misleading, it’s harder than ever for anyone to trust any particular one. When trust is low, our heuristic for truth becomes: ‘does this line up with things I already believe are true?’. And then we’re surprised when our fancy new infographic does a terrible job of challenging preconceptions.
And no, it’s really not enough to say ‘I referenced my sources’. Academic referencing is, quite simply, shitty interface design, and we can do better. In a world of decreasing attention span, we need to do better.
So what does this new generation of representations look like? To be perfectly honest, I’m not sure yet — it’s not an easy problem. But one thing is clear: they need to invite questioning, rather than ignore it. One way to approach this could be the “5 why’s” test. That is, a viewer can ask “why” of any (reasonable) part of the representation and get an answer, then ask “why” about the answer, and so on, until they hit 5.
Right now, I’d be surprised to find anything that properly hits 1 ‘why’.
To achieve this, they’ll have to use some combination of the tools used in interaction design, data visualisation and game design. They’ll be harder to make and take longer to build. They’ll probably have to build on each other, to reduce the terrifying amount of work. The flexibility and re-usability of the format is also important. Possibly, they’ll have to incorporate actual user input where the designers can evolve the representation past the point of initial release. There are a lot of possibilities here.
Suffice to say, there’s a lot of work to do. If you’ve got any thoughts on this, or you want to tackle this problem, please get in touch.
Something I’m hearing a lot is that a failure to listen is at the heart of many of the divisions that we see in the current political landscape. If this is the case, isn’t it time that our representations of data started to listen?