Expert views on data literacy and alternative forms of data visualisation
As a designer I see great potential in visually combining qualitative and quantitative data, and thus in teaching visual thinkers to become more data centred. But how do others see the potential and limitations of this project? In order to shed light on the topic from a different perspective, I talked with four experts in data visualisation: Peter Lovei (Data Scientist) and Erwin Hoogerwoord (Data Visualiser) at Philips, Jacky Bourgeois (Assistant Professor) at TU Delft and Eva Ruitenbeek (Data Designer) at Eyeon. Topics we discussed included: teaching data literacy, potential and limitations of the tools I’m developing, and the balance between accuracy and creative freedom.
When asked about their current practice, the experts agree that the combination of quantitative and qualitative data is something they do mostly through annotation. Another mention was the method of quantifying data, using tools such as text mining or machine learning. Overall the techniques are focussed on creating a holistic picture. Peter Lovei explains this important aspect of combining data:
“The trick is to create a balance between the bigger picture storyline, and nuances in the data using specific points or quotes”.
When going further into the communication aspect of data visualisation, the usage varies from a quick discussion tool in interviews to the reporting of results or predictions. A vital aspect in this is knowing your audience and their relation to data, as Jacky Bourgeois points out:
“Interviewee’s, when you’ve done your job properly, are the experts of their own data. It takes much less time explaining the data to them, than to researchers who are unfamiliar with the data”.
Pros and Cons of new visualisation techniques.
In the research project, I am using different illustrative techniques to integrate datasets in visual consultancy. We discussed the potential of these new tools and processes, as well as its limitations. Overall the new, integraded data visualisations are perceived as clear and easy to read. Erwin Hoogerwoord mentions that the information density of a visual is similar, if not higher than a traditional dashboard because the images are not limited to bounding boxes. Eva Ruitenbeek adds to this that there is more freedom in designing the visuals in this way, as software tools have the limitation of screens and fixed dimensions which makes some graphs (especially with many data points) harder to read. Jacky welcomes the use of metaphors:
“A big thing I have learned from data visualisation is that it is about storytelling, not only about what is, but how we perceive it. Especially for dynamic elements a common metaphor from outside can be very useful”.
Of course there are some limitations to the current visualisation method, such as a low level of interactivity possibilities, and the possible misinterpretation of 3D visualisations. More importantly the danger lies in accepting client’s data too quickly. So how do we teach designers and visual thinkers to become fluent in data literacy and translating this to visual communication? Especially on this topic, the experts had loads of input, of which the most interesting ones I will share with you here.
Advice from the professionals
Asking the right questions. By far the most important advice is that we need to start the data thinking process in the early stages of the consultancy process. The early sessions give opportunity to ask the right questions about the data that was collected and in which context. Erwin stresses the value of using consultancy and research to make sense of the data first, rather than cutting and pasting them directly in a visualisation.
"As consultant, build up confidence to ask questions when receiving the data. You could, for instance, plan a meeting in which the client explains their own data as a platform for discussion."
Peter stresses the importance of annotation and/or co-creation, as this allows different parties to make sense of the story together. Creating stories together is something we are very good at as visual consultants, but is not yet used in a way where we use a client's data as starting point.
Process steps. In order to create the right impact, a few aspects need careful consideration. Eva explains three important process steps:
"Reduce the amount of information to the core insights, which should be honest and correct. The core data should add to the goal of the project, to help you further in some way. Lastly, make sure the data visualisation adds to the readability of the image, rather than increasing clutter."
A good addition to this is creating a dynamic storyline, which can navigate the reader through the data. This increases the readability, and your control over the way in which viewers gather knowledge.
Catering to the right audience. During the design, consider the audience and goal of the project. Not only as Jacky mentioned before, to sufficiently explain the context and data to outsiders, but to find a balance between honesty and catering to the right interpretation. Therefore, I'd like to close with this:
"We have a huge responsibility in correctly showing information. This is a big discussion topic currently. It is important that even when the data is lacking, your interpretation doesn't make it worse"
"Make sure it is presented, don’t let data or interpretation fill possible gaps. Present what is unknown, to show what is yet missing”.
A big thank you. These interviews were incredibly valuable, both in validating assumptions and giving new insights from the data perspective. I’d very much like to thank the participating experts for their time. For future articles I am still looking for those on the other end of the spectrum: want to make use of data to support your communication, but have no idea what or how? Let’s talk!