Data + Science

[ad_1]


12/4/2019
Text Alignment on Dashboards


Yesterday, Steve Wexler posted a new blog post DON’T center, right-align, or justify text on a dashboard. Steve makes a case that only left-aligned text should be used on a dashboard. While Steve states in his blog post, “Or at least have a very good reason for doing so.”, he doesn’t include many examples of this. I felt compelled to write this blog post to offer some additional examples where centering, right-aligning or fully justifying text does not interfere with the interpretion of the data, and even an instance or two where it’s critical that we SHOULD do this.

Note – Steve modified his original blog post to clarify his concerns are about what happens when you have multiple lines of text and not single-line items.

Let’s start with the title of our visualization. Steve, and many others, advocate for setting the title in the top left corner of the visualization. As he and many others point out, the top left corner is the natural starting point for where our eyes begin when reading a visualization and we read from left to right in a z-pattern (at least in the western world). I don’t disagree with any of this. This is solid adivce. In fact, this is my default placement for my chart titles. Have a look at my Tableau Public Profile page and you will notice that almost all of my visualizations have the title left-aligned in the top left corner of the visualization. If using Tableau Server or Tableau Online, then one additional advantage of left-aligning the titles is that they will show up consistently on the thumbnail image. If the title on the visualization is centered, then you risk cutting it off in the thumbnail image.

That said, you will also notice on my Tableau Public Profile page that there are a number of instances where the title is not left-aligned, for a variety of different reasons. When choosing the position of the title, I also consider other elements of the visualization. For example, the visual balance, the grid design, the focal point and how much text is being shown. Here’s one example.

In my Beautiful Trash visualization, almost all of the text on the visualization violates Steve rule, “DON’T center, right-align, or justify text”.

  1. The title “BEAUTIFUL TRASH” is centered on the visualization to create balance.
  2. The size of the subtitle text is set specifically to create a fully justified text look.
  3. The bar chart labels for the day of the week and the miles driven are center aligned above and below the bar. This is very common.
  4. The label in the annotation on the map is center-aligned.

These designs about the text alignment create a visualization that is centered. While this example is not a typical business dashboard, it showcases a number of instances where it would ok, if not preferable, to change the alignment of text and not left-align all of the text.

Let’s look at one of Steve’s dashboards from The Big Book of Dashboards.

Notice in this example that Steve centers the four categories of the dot plot with jitter, as well as the label in the center of the selected dot. This creates a beautifully centered look on this dashboard. Could he have left-aligned the category labels? Sure, but since the selected mark is centered it makes sense that the category labels and dot labels are centered. Does the centered text make it harder to read? No, because it’s just a few words of text that are very easy for anyone to read quickly.

Here is another example from Five Thirty Eight where they visualize the Eurovision ranking using a slopegraph.

Notice that on the left side of the slopegraph they right-align the text labels and on the right side of the slopegraph they left-align the text labels. This creates a very nice balance in the visualization, with the slopegraph acting as the center point for all of the text.

The example that sparked Steve’s blog post was Breaking the Barriers in Baseball by Chantilly Jaggernauth.

Chantilly talks about text alignment in her presentation Design Secrets for a Non-Designer from the 2019 Tableau Conference. What stands out to me in this design is the beautiful balance between the left side of her visualization where she is right-aligning the text and the right side of the visualization where she is left-aligning the text. I think this design decision works very well here. You might be thinking, “sure, this works well in this case, but most business dashboards don’t have this type of alignment down the center.” Agreed, but I’ll counter that most business dashboards don’t have paragraphs of text either. And while reading a paragraph of right-aligned text may be more cumbersome, reading a word or two is really not much trouble at all. So let’s look at some additional examples that aren’t using paragraphs of text.

One example where right-alignment of text is often used is for bar chart labels. Here is an example from my Complaints Dashboard featured in The Big Book of Dashboards.

Notice the right-aligned labels on the bar chart.

It is very common to right-align the labels on bar charts. Stephen Few does this on most of his bar charts and bullet graphs (examples here and here). Cole Nussbaumer Knaflic, author of Storytelling with Data, right-aligns the bar chart labels here and right-aligns text on the top right hand side of the chart here (Figure 6.3 from her book Storytelling with Data). The Urban Institute Data Visualization Style Guide, which is an excellent style guide, specifically states “When using a horizontal bar chart, right-align the category labels and center them vertically with the respect to the bar.” Steve Wexler even uses this technique on a number of his visualizations, most recently in his Comet Chart. Let’s look at another example where it is critical to right-align the text on a dashboard.

“DO right-align text on a dashboard”

There is at least one example where it is very important to right-align text on a dashboard; when using a table of numbers. I have written about this before in my blog post 5 Tips on Designing Text Tables. In most table designs, it is important to right-align the numbers so that the users can easily scan the table for difference in magnitude. Here is an example of a highlight table from the Churn dashboard featured in The Big Book of Dashboards.

Notice how easy it is to see the difference between two digit numbers (ex. 60 and 70) vs. the three digit numbers. This is easy because the numbers in the table are right-aligned.

In addition to text tables, BANs would be very similar. These big numbers are often right-aligned or center aligned and in either case it would make little sense to left-align a text label if the BAN were right-aligned or center-aligned.

Here is an example of center-aligned BANs and labels from my Complaints Dashboard featured in The Big Book of Dashboards.

And this example of center-aligned BANs and labels from the Churn dashboard featured in The Big Book of Dashboards.

This is not to say that BANs and their labels cannot be left-aligned. Here is an example from my London Buses Safety Dashboard where the BANs and labels are left-aligned.

Overall, Steve Wexler’s advice for aligning text is genrally good advice. However, as with many of the “rules” in data visualization, we have to consider the many caveats. It is often difficult to create blanket rules to “do this” and “don’t do that”. As explained in this post, there will be many instances where using a different alignment for the text will create a more balanced visualization or better grid design. In the case of the text table or a highlight table, it would actually be best practice to right-align the text. One of the reasons I felt compelled to write this post was that I was worried that people might read Steve Wexler’s post and apply the “all my text should be left-aligned” rule to their text table. My advice is to think carefully about all of the elements in the design of your dashboards and other data visualization. While hard and fast rules can be useful, especially to people just starting out, these rules can often be applied too broadly.

I hope you find this information useful. If you have any questions feel free to email me at Jeff@DataPlusScience.com

Jeffrey A. Shaffer

Follow on Twitter @HighVizAbility



Read More …

[ad_2]


Write a comment