On February 13, 2023, we had this question proposed to us at the beginning of the day, So how do we make a Viz more effective, or simply, easier to digest?
Let's start simple, identify the general idea or topic the chart is displaying.
- This information can be found mainly in the title, the variables on the axis, and the legend (if it's there).
- Here we have a 3D bar chart
- The y-axis are the percentage, x-axis are the years
- The bar represents population differences
Next, once we identify what the chart is trying to say, we can start cleaning up the junk that clutters the chart.
- Chart Junk, is a term for no substance design as it doesn't add any purpose to what the chart is trying to tell and often distract the viewer from the meaning.
- In this step, we will remove those unnecessary design elements and reorganize the design.
- This could also mean using a different graph that is better suited for the data displayed.
- The point of this is to make the graph have a more consistent display.
- Some of the things to consider
- Can the title be simpler? Is the title directly explaining the chart?
- Does the graph have any consistency between the Fonts? Colors? Size? Etc
- First, I remove the 3D effect as it makes viewing the data difficult.
- Return it into a 2D bar graph.
- For a better comparison between the 2 populations, I turn it into a line graph to show the population difference, approximately when does the population intersect, and hold a clear view of how they are compared.
- Line Charts are usually used to show the approximate value between 2 data.
- Here I did a bit more cleaning
- Change the title, shorter and much more direct
- Shift the start of the line graph onto the y-axis
- Remove the legend and label the lines
- Remove the gridlines for a more clear view
Now that we have finished cleaning up the graph, the next step is to see what can and should be added to the graph without cluttering it.
- One thing to consider is whether the context is all there.
- Does anything need to be labeled?
- What should be highlighted?
- Any annotation that needs to be made?
- How does the alignment of the graph look?
- Does it match up? Does it look funny?
- Do you need to add a filter to the graph?
- Should your viewer have any freedom in navigating the graph?
- Sometimes the graph just needs some cleaning, so there is no need to add anything else.
- Added the Year label to the x-axis and the population percentage in the y-axis
- Kept it out of the way
Now that we have added context to the graph, we are almost done. The last step is to ask yourself “Can I understand this? Can my audience understand this?”. If you are hesitant to answer “Yes” or if it’s a “No”, start from step one and clean up the work.
One thing we have to keep in mind is, what can be changed. This is separated between Data Ink and Non-Data Ink. Non-Data Ink is generally the representation of the measurement, generally located on the axis. Can not be changed. However, for Data Ink, it represents the design of the variable, such as color, size, etc. Understanding the difference between Data Ink and Non-Data Ink is the first step in creating an effective Data Viz.