“Perfection is achieved not when there is nothing more to add, but when there is nothing left to take away”
(Antoine de Saint-Exupery)
Recently, I got an opportunity to work on a data visualisation project. Initially it appeared to be a simple graph but when I went deep, there were a lot of factors involved to make it more insightful. I have not shared the screens here due to confidentiality of the project but I have documented important points to kickstart the visualisation.
Fundamental Goal: Communication
Parameters to be considered before designing (From Stephen Few’s ‘Information Dashboard Design’ book) :
Update Frequency Daily
User Expertise Novice
Audience Size One person
Multiple people with the same requirements
Multiple people who need to monitor different data subsets
Technology Platform Desktop/laptop
Web server/ browser
Screen Type Extra-large screen
Data Type Quantitative
Once you are clear with the above parameters, it is time to go into design mode. Below are few points that can help in designing effectively.
1. Choose the right display media: For displaying the right chart, the most important thing is to understand the need. Eg: To show comparisons, Bar charts would be more effective and to show pattern over a period of time, line graph would be a better choice. Here is the reference (by A. Abela) for the same.
2. Identify the Content that really matters:
#Rule 1 Data should be the star of the show.
#Rule 2 Add meaningful context.
#Rule3 Go by ‘Less is More’ approach. Maximize data-ink ratio.
Edward R. Tufte offers this design principle:
Maximize the data-ink ratio, within reason. Every bit of ink on a graphic requires a reason. And nearly always that reason should be that the ink presents new information.
Below is a nicely illustrated example by darkhorseanalytics
3. Use Color judiciously:
#Rule 1 Use color only when needed to serve a particular communication goal.
#Rule 2 Use soft, natural colors to display most information and bright and/or dark colors to highlight information that requires greater attention.
#Rule 3 To guarantee that most people who are colorblind can distinguish groups of data that are color coded, avoid using a combination of red and green in the same display.
This is an amazing newsletter by Stephen Few on the use of colors in charts.
Also you can use this tool to checkout the appropriate color combinations for the charts.
4. Define a suitable Layout:
#Rule 1: Keep flexible layout to produce a dense information display that a viewer can easily and rapidly monitor.
#Rule 2: Items that relate to one another should be positioned close to one another.
#Rule 3: Identify the right aspect ratio and let the optimal ratio define the layout. Eg: the additional information can be put to the side of the chart.
5. Add simple Interactivity:
#Rule 1: Choose filters wisely. The timeframe to show the data should be chosen as per the project need.
#Rule 2: Put supplementary information within reach. Eg: In order to view deeper level of detail, a pop-up box (tool-tips) can be shown on hover.
Hope you like the post. Do share your thoughts 🙂