Little Crash-Course in Big Data – Visualising Data
A designer’s guide to quantitative data from one amateur to another (Part 3 of 3)
This article is the third and last, in a series of three about how to collect, analyse and visualise quantitative data targeted to designers who want to get into quantitate data. You can find the other parts here - Collecting Data and Analysing Data.
In terms of dealing with data, visualising it is probably where the designer feels most at home. Much has been written on the subject but all the same, I thought I’d share a few cautions.
Visualising the data
When visualising data, it’s important to think about how to emphasise the key points that you are trying to convey to the audience with this particular visualisation. The visualisation should show the underlying data without distortion, and avoid common pitfalls that obscure the real information.
Below is a good guide to choosing an appropriate visualisation method based on whether you want to show a comparison, a relationship, a distribution or a composition.
A successful design usually has to be visually appealing but people who have worked with data a lot longer than us, have thought about this before. Instead of shooting from the hip and choosing a sexy pie graph, learn from the experts.
PIES ANS CIRCLES
Pies charts and bubble charts are not well suited for accurate or precise comparisons. This is because you’re asking the viewer to compare two-dimensional areas rather than one-dimensional graphs and this is very hard.
Ironically, these two methods of visualising data are very popular in the design world. In addition, some people (and different software) graphs bubbles by volume, not diameter, further complicating comparisons.
To illustrate the problem, have a look at the two dark circles in the middle. They are the same size but context influences how you see them.
An even more powerful illustration is the example below. It’s hard to see the difference between the three smaller pieces of the pie chart. Comparing them in the more appropriate bar chart is no problem at all.
VISUALISATION HELPS
In the example below, the four scatter plots A-D have the same mean, variance and correlation (the dotted line). Yet when we look at the visualisation of the distribution we can see that they are very different. Read more about this example here.
The example above is often used as an example of why it’s so important to visualise data. What is more worrying however, is that visualising can significantly misrepresent the data unless you’re aware of things like significance (p-value) and how to appropriately visualise data.
LIMITATIONS OF VISUALISATION
Visualising is a great way to communicate your results but there are limits to what the naked eye can see. Visualisations don’t tell the whole story.
The naked eye and unaided brain can’t easily compute large amount of quantitive information. Have a look at the more realistic scatter plot below. Without the line, it’d be hard to see the non-linear correlation.
It’s easy to put too much stock in visualised data but it can be misleading as we learned in the second article in this series about Analysing Data. Bar B looks convincingly bigger than A but is the difference between the two that significant?
Conclusion
Ok, you caught me. This was not a crash course about Big Data rather a 101 on Plain Old Data. I don’t really have the depth of knowledge to teach you much about Big Data but I hope you learned something.
I will finish with a passage from one of my favourite books, Equal Rites by Terry Pratchett. Cutangle and Treatle, two university professors are conversing after a mind-blowing lecture at the Unseen University.
“Cutangle: While I’m still confused and uncertain, it’s on a much higher plane, d’you see, and at least I know I’m bewildered about the really fundamental and important facts of the universe.
Treatle: I hadn’t looked at it like that, but you’re absolutely right. He’s really pushed back the boundaries of ignorance.
They both savoured the strange warm glow of being much more ignorant than ordinary people, who were only ignorant of ordinary things.”
Part 1: Collecting Data
Part 2: Analysing Data
Part 3: Visualising Data (this article)
A bit about me
My name is Stina Jonsson and I work at IDEO in our London office. I apply cognitive science to design challenges to fundamentally question and reframe the way we engage in a digital context.
PS. Sorry for the spelling mistakes.
