Visualize This

Dear Washington Post,

Making a good infographic is hard. You have to begin with a comprehensive, consistent, valid set of data (the info) and rely on skilled design (the graphic) to illustrate whatever narratives and insights—which are often either subtle or seemingly contradictory—are hiding in the numbers. People who do it well get paid a lot of money. The firm I work for makes them, so I know first-hand. You’ve done it well yourself from time to time, so I know you understand, too.

Speaking as a sympathetic friend, then, I really have to tell you: your snapshot of major theater venues in the Washington region is more than a bit of a mess.

Let’s look at your data first. Is it comprehensive, consistent, and valid? I’ll give you one out of three. You’ve clearly done your research (I’d expect nothing less), so the data you do have seems to be accurate.

But is it comprehensive? Hardly. You’re missing easily sixty other institutions that ought to be included in any reasonable analysis, if not more than that. I don’t expect you to capture them ALL, mind you… but you really ought to get a lot closer to 100% if you want to be reflective of the entire region. (Don’t have the resources to be fully comprehensive? Then at least pick a more representative sample. You’ve captured only the largest institutions in the region.)

And is it consistent? Your infographic’s title suggests you’re comparing venues, but a few of the institutions on your list aren’t venues at all. You’ve blurred the line between venues and theater companies, which is really problematic.

Next, let’s look at your graphics. You’ve got a timeline that indicates both the year each institution was established and its size, and you’ve got a chart that lists a variety of information about each institution.

The timeline I find particularly baffling. What, if anything, were you trying to reveal by comparing size to year of establishment? There’s nothing particularly enlightening about this element of the graphic at all: no trends, no reversals of trends, no correlations—nothing. If this is a “snapshot,” Washington Post, it’s a really awkwardly-posed portrait, I must say.

The chart, by contrast, suffers from a major apples-and-oranges problem. Alongside several raw data points, you’ve listed what I can only call “impressions”—the changes that have been made in each institution’s recent history and the nature of their programming. The information is highly subjective in both cases, which undermines the legitimacy of the infographic (and, in addition, dilutes the data).

The chart also includes one calculated data point: the percentage of each institution’s budget that comes from government funding. Finally, I would say, a potentially interesting narrative buried in the chart. What I wish is that you’d done a great deal more of this.

For instance, using only the data you’ve got in your chart, you could have calculated the percentage of each institution’s budget that goes to the top-salaried worker’s income. You could have compared average attendance to the number of full-time employees, too, to try to find a correlation there. Those are missed opportunities.

But what’s really disappointing is the data you didn’t collect: the number of tickets sold by each institution in the previous 12 months; the numbers of unpaid interns (or staff) supporting each institution; the salaries of the lowest-paid workers at each institution. You could have probably done a lot more. Think of what your readers might have learned!

Here’s just one idea. For a good while now, there’s been a healthy dialogue about the ratio between the highest-paid workers at arts institutions and the lowest-paid workers. Would that have been hard to dig up? Perhaps it would have… but I can promise you that your readers would have found that ratio highly compelling. (Well… those who aren’t in positions of power at big theater institutions, anyway.) And isn’t serving your readers what you’re really trying to do?

Yours in data visualization,

Gwydion