Thirteenth Thing: Visualising Data
Visualising information with colour can help our target audience make sense of research data or information. Colour can draw attention to key details and encode dimensions of the data. The great thing about using visual channels to communicate research is that we can reach diverse audiences, especially when designed to cross language or vocabulary barriers – Amanda Belton and Usha Nattala take you through how to visualise your research in today’s Thing.
Different ways of visualising research with colour
One simple step can be to create a colour dictionary for your graphs, infographics, maps and conceptual model diagrams. Rachel Diprose and team used colour encoding for the themes in their case studies and these colours were consistent across the research website and their other publications. If you have an image already that fits with your intent, Adobe’s online tool will give you an appealing colour palette that you can screenshot or download. A screenshot is simple and if your favourite design tool has an ‘eyedropper’ function you can use it to ensure the colours are consistent across your figures. Keeping to a consistent colour palette is more than aesthetics, it builds a mental map in our audiences’ minds that helps find information when scanning so viewers can fit new information into an existing mental model, reducing the cognitive load.
These animated gifs were designed to invite people into Carmel Mesiti and team’s research on the lexicon of terms teachers use in their professional practice. The colour green representing the Finnish lexicon, and pink for the Japanese lexicon. This design aims to use colour to signify the language of the lexicon.
Every dataset has a story to tell: a spatial dataset often needs a map to bring out the full intensity of what it is trying to convey. Maps have strong conventions and key considerations to display geo-spatially organised data. The conventions are the usage of certain colours for certain landscape features such as green for vegetation, blue for water and bold outlines for place boundaries like in Melbourne Pollen and University of Melbourne Seismic monitoring. Key considerations are spatial arrangement, balance and positioning of coloured shapes, scale and crowding. A good map honours the place it is meant to showcase and is an artwork in own’s right.
Colour and accessibility
Colour is powerful and we know with great power comes great responsibility. Checking any visual outputs in greyscale can help keep the design accessible to people with colour blindness. For example, Microsoft Powerpoint has a ‘colour saturation’ feature to check if an image is still readable in monochrome. There is a limit to the differences we can detect in shades of colour: if you’re building your own colour dictionary then Brewer’s Palettes will suggest a distinguishable colour scheme for the type and scale of data dimensions you want to visualise. Colour encoding can help ‘read’ information, but insufficient contrast between your coloured background and your font colour, means the colour is readable but the text is not.
Likewise, often we use plain grayscale maps as a background for strongly coloured foreground information to minimize the noise and make it more accessible. Even when we use coloured maps, we keep the background colours and text muted to provide sufficient contrast. Where a palette of foreground colours is needed to visualize the range of some data features, such as say the percentage of young people in each local government area, it is important to ensure that the scale of colours used have a sufficient contrast between themselves, like in Circuit by Rachel Fensham and team. The human eye perceives more colours in the green spectrum but cannot distinguish between shades of red very well. Thus, we can use more shades of green but must be careful with red. The four colour theory states that you need a minimum of 4 colours to fill in shapes on a map such that no two adjacent shapes have the same colour. To choose such a palette with 4+ contrasting shades of a colour, we often make use of colour brewer tools for maps.
Using colour to communicate meaning
The power of colour to communicate meaning can depend on conventions and culture. When choosing colours we need to consider the context. You may have seen Ed Hawkin’s warming stripes showing changes in global temperatures. Blue for cool temperatures and red for warm is a common, conventional colour association with temperature data appropriate for global information. Whereas, Kontinentalist, a design studio, use colour conventions tailored to Asian countries. When Liz Dean collaborated with MDAP’s Mar Quiroga on the design for a research site on Decoloniality and Thinkers and Practice, they chose colours to reflect the different nations for each thinker to act as visual wayfinding that can be quickly processed by our vision when navigating through the resources.
About the authors
Amanda Belton is a research data specialist, at Melbourne Data Analytics Platform, working with education and arts researchers to visualise data. She works with playful approaches and empathetic design principles to communicate research data visually.
Usha Nattala is a research data specialist specializing in machine learning, timeseries analysis, app development, geospatial analytics, data science, data modelling and visualization, and math-intensive programming.
Interview with Alana Pirrone
What is your role about?
I’m the Design and Communications Coordinator for the Child and Community Wellbeing Unit in the School of Population and Global Health. My role involves translating public health research outcomes into visual designs to aid in comprehension and create maximum impact. As part of this role, I also run a consultancy service to upskill researchers in design and data visualisation best practice. I have a number of short courses, seminars and workshops which explore how to visualise your research and use data storytelling to communicate your message to your desired audience.
How have you used/interacted with data visualisations in your research/role?
I train and upskill researchers to be able to use data visualisation effectively to translate and communicate their research. This can include choosing the right chart to display their data (i.e using a slope chart rather than a bar chart to display changes over time), all the way to developing a graphic that captures and visualises their qualitative findings. Data visualisations and graphics can then be tailored and adapted to specific mediums. For example, a chart that you would design for a report or infographic, may not be the best chart to show in your PowerPoint presentation, as people have less time to read and take in the information, especially if the speaker is talking over the top. Keeping your desired audience at the forefront of your mind when designing is key.
How have data visualisations helped you work smarter, not harder when managing your research?
Data visualisation in an effective tool in communicating your data to enhance comprehension and retention. People are busy and don’t have time to read through long reports or try to decipher clucky charts. Data visualisation is an effective tool in communicating your research or data efficiently and succinctly.
What is your number one (top) tip for visualising data?
Focus on the message you are trying to communicate to your audience. What do you want them to know or do with your data or your research? What are the interesting parts? Highlight it using visual attributes like colour, contrast, size and weight. Draw emphasis to that part of the chart. Use a full sentence as a heading to describe the chart and the story that you are trying to tell.
Alana Pirrone is a Design and Data Visualisation Consultant. She specialises in data visualisation for knowledge translation. Alana has particular experience in translating public health research outcomes into visual designs, and has been commissioned by the World Health Organisation, independent consulting firms, community health services and non-for-profits.
Cite this Thing
You are free to use and reuse the content on this post with attribution to the authors. The citation for this Thing is:
Belton, Amanda; NATTALA, USHA; PIRRONE-SAVONA, ALANA (2024). Thirteenth Thing: Visualising Data. The University of Melbourne. Online resource. https://doi.org/10.26188/25340296
Categories
Leave a Reply