# Depth and breadth in charts

“Explore your information at multiple levels of depth and breadth,” data vis designer & journalist Alberto Cairo writes in his book “The Truthful Art” (p94). The Data Vis Book Club discussed this book yesterday. So for this week’s Weekly Chart, it’s just appropriate to design a chart that takes his statement by heart:

We can read a few things out of this chart. First, we learn how much energy different countries consume[1]. But this information is mostly there to compare it to the energy consumption of the IT sector worldwide. To explain what this sector contains, it’s subdivided into four parts:

• Devices: How much electricity do our tech gadgets need when we use them (e.g. laptops, TVs, smartphones, gaming consoles or printers)?
• Manufacturing: How much energy is needed to produce these tech devices? Corcoran & Andrae, who estimated the data in the chart, note that “for most
consumer electronic devices the electricity that is used in the manufacturing process is often
as large as the lifetime operational electricity of the device. In some cases, it can be
multiples of the operational energy.”
Well, wow.
• Networks: How much energy is needed to let our devices talk to each other? This includes the energy used by base transceiver stations, switches and routers.
• And Data Centers: How much energy do servers need to store the selfies we message and the emails we write? That’s what we talk about when we talk about the cloud.

### Chart Choices

Let’s go back to the first idea: “Explore your information at multiple levels of depth and breadth.”

First, the depth. Alberto didn’t name his book “The Truthful Art” for nothing: He talks a lot about how we can make our charts truer. To do so, we should think about how to reveal the complexity of the data and the uncertainty in it, e.g. with error bars.

It’s inconvenient to add error bars to a stacked bar chart – but we can just add multiple bars. We communicate two important ideas when we show the three estimates the researchers calculated instead of just the estimate for the “expected case”:

Second, how uncertain we are. The IT sector in 2015 might have consumed two times as much energy as Russia, or three times. We don’t know: The range between estimates is huge. (Which makes it even more important to show them!)

So we add depth through visualizing the different estimates and the subparts of the IT sector. Another option to add depth is to show the data over time or for different countries.

And the breadth? We can introduce it by setting our estimates in a bigger context: How much are 2000 to 2500 terawatt hours of energy consumption, really? Turns out: a lot. Country-level-lot. Big-country-level-lot. Only now we can imagine what so much electricity means.

Adding both breadth and depth to our data will make our charts more useful to our readers. As always, you can hover over the chart above and click on “Edit this chart” in the top-right to see how I created it. Also, follow @datavisclub if you want to hear about other data vis books we’ll read. I’ll see you next week!

1. Wikipedia has you covered in case you’re looking for energy consumption per capita. ↩︎