Viz Wiz 101

Whenever I think of graphic design and psychology, Gestalt principles is a term I frequently hear when human perception is involved. Having heard this term many times before, Gestalt principles is not something I’m unfamiliar with. I find design principles and the impact on human perception constantly creeping up in anything related to visual design. Prior to taking an in-depth data visualization 101 class at the Data school, I half heartedly thought utilizing some of the principles like similarity or proximity was enough to communicate my data effectively without considering the holistic message of my data story and the actual practices that would best express the message to the audience. Ultimately, it’s about letting the data do its job in letting the reader explore your dashboard. The best practices we learned in class and the warnings we received about chart junk are really just pushing us to consider the data-ink ratio through removing elements of our data visualization that don’t add anything to communicating the data.

I really began to understand data visualization through Gestalt principles and the pre-attentive attributes. When you visualize length as a bar graph, orientation as slope, scatterplot as position, or pictogram as shapes and think about the best practices that optimize the shape of each type of visualization, it becomes easier to think of what visualization can work with your data.

Thinking more about the data ink ratio, I learned that many bad habits can take form in chart junk. The chart junk I briefly mentioned in the first paragraph are graphics that generate redundant data ink that doesn’t add anything to the visualization. Chart junk makes it harder for the audience to understand the data and can often distract the data exploration process for your viewer. Chartjunk can include:

  • Heavy gridlines
  • Unnecessary text
  • Unnecessary color
  • 3D Effects
  • Icons
  • Pictures
  • Axis Ticks
  • Redundant Labels
  • Redundant Titles
  • Shading and Gradients

When I was thinking of clean design and implementing the best visualization practices, I remembered we mentioned in class that bar graphs and line graphs are the most commonly used and easily understood. Using some of the best practices for data visualizations, I wanted to create two examples of a good and bad line graph to create a side by side comparison:

Some needed changes include removing the gridlines, changing the line colors to more contrasting colors, reducing the length of the title, and making the annotation more relevant to the line graph. The colors chosen for the widgets are too similar and the gridlines are too distracting, which takes away the audience’s attention on the data. Relating to concepts of the graphs, the title is not short and succinct and the annotation doesn’t say anything about the data. If I were to recreate the data, it would look like this:

After removing the gridlines and changing the colors of the widgets to a more contrasting color set, the data can be seen more clearly. The title is straight to the point to retain the viewer’s attention and the annotation adds context to the data rather than take up space. Hopefully this side to side comparisons will help you create cleaner data visualizations that puts the data exploration first. In honoring the data ink ratio, we’re allowing the best usages for every visualization to come out.


Author:
Connie Koo
Powered by The Information Lab
1st Floor, 25 Watling Street, London, EC4M 9BR
Subscribe
to our Newsletter
Get the lastest news about The Data School and application tips
Subscribe now
© 2025 The Information Lab