The advent of ‘digital’ introduced a swathe of implications. Brands are now able to access customer sentiments in real-time and respond swiftly – in effect, have meaningful dialogues with the people that matter. Beyond this is its power to produce data that enables organisations to understand what consumers like or don’t like, what they are interested in, and interpret optimal messaging and communication formats.
With a caveat: data is meaningless without the ability to know what data points are important, how they need to be analysed, and why it’s contextually important. This all comes down to data literacy.
What is data literacy?
Data literacy is the ability to read, write, and communicate data in context – it is the ability to derive meaningful information from data, and communicate that meaning to others. It’s a skill that empowers people to ask the right questions, build knowledge, make decisions, and communicate meaning to others, using data.
Therefore, data literacy is set to become as fundamental as reading – “it can change the world in the same way that literacy transformed access to knowledge in the centuries after the printing press. Instead of trusting someone else’s interpretation of what the data means, we’ll be able to go straight to the source…Instead of making decisions based on passionate opinion, we’ll value curiosity and intellectual honesty”.
Being data literate means being able to understand data. Data can be presented in a number of different forms – for example, a bar chart, pie chart, a radar, or even a table. The user or audience needs to be able to understand what is presented in the visualisation – what does the data tell us? What insights can be derived from it?
However, this is only part of the story – “Just because data is presented in a pretty chart, it doesn’t automatically mean that it’s easy to comprehend or that it should be trusted”. Data literacy and understanding data also requires thinking critically about what you are presented with – this means considering the quality of the data, the analysis, and the integrity of the visualisation. Visual integrity – or graphical integrity, as it’s also called – means ensuring that what’s presented in visualisations accurately represents what’s in the data, and that no design choices distort the findings.
Data visualisation
If you are interested in data on interest rates, visualising the data in a bar chart is an effective and simple way to get a good overview of changes in rates over the years (as in the image below). However, if an inappropriate scale is used for the vertical-axis (Y-axis), the visualisation can imply that there is a large increase of the interest rate, when in fact, the actual difference between years is minimal. Being data literate means not only being able to read and understand the chart on the right, but also understand why the chart below is problematic.
Not all cases of charts lacking visual integrity are innocent mistakes – sometimes visualisations intentionally distort the data to tell a particular story for a specific reason. The power of data and data visualisation is finally being realised and both are being used to support and inform important business decisions. To be able to critically evaluate data and turn it into effective data visualisations and stories is fast becoming one of the most important skills.
Data visualisation examples
Let’s take a look at some examples of data visualisations designed to purposefully mislead and support a particular agenda. In the visualisations below, the Georgia State Public health Department intended to make it look like COVID-19 infection cases had decreased since lockdown was lifted. As you can see once the chart has been correctly organised, this was not the case. Even when we reorganise the chart, there are still issues of consistency that make comparisons hard – the counties (represented by different colours) are not listed in the same order for each date, for example.
A misleading chart notice how the variables along the horizontal-axis (dates) are not ordered (source)
Another example, also from the USA’s COVID-19 reporting, clearly misrepresents data by using an inconsistent scale on the vertical-axis. Looking at the image below, it’s easy to see how the inconsistent scale distorts the information and knowledge communicated by the visualisation, and how this can lead to actions grounded in false wisdom.
Note how the inconsistent scale on the vertical axis distorts the information and knowledge (source)
As we have seen, visual integrity highlights important concepts around data, data visualisation, critical thinking, and ethics. Michael Correll notes that the power of data visualisations to influence decisions means that designers and analysts have moral obligations and ethical duties to ensure the integrity of their graphics. It also means that being data literate is becoming a necessary part of critical thinking.
In our effort to ensure humans have the critical skills to stay ahead, we partnered with industry experts to create an original and up-to-date course on data visualisation. This course will help you build the knowledge and skills you need to be data literate and most importantly present the right ‘story’. Ben Jones has identified 17 key traits of data literate people, based on his years of work as a data professional. He has grouped these into four categories: knowledge; skills; attitudes; and behaviours. Our course will build your knowledge and skills in data visualisation, and encourage the attitudes and behaviours common to data literate professionals. Sign up and get started!
Author: Gareth McAlister, Senior Instructional Designer at Red & Yellow. Gareth loves data! His Masters in Anthropology has seen him involved in a number of data-related endeavours, from exploring the use of non-timber forest products to mineral, soil, and botanical diversity sampling, to sentiment analysis, marketing and web analytics, and education. He is passionate about data literacy, gamification, and online education. He can usually be found rolling dice and running a game of D&D for friends.
