Small tweaks that make big differences

It's one of those days that a web search led me to an unfamiliar corner, and I found myself poring over a pile of column charts that look like this:

This pair of charts appears to be canonical in a type of genetics analysis. I'll focus on the column chart up top.

The chart plots a variety of gene functions along the horizontal axis. These functions are classified into three broad categories, indicated using axis annotation.

What are some small tweaks that readers will enjoy?

***

First, use colors. Here is an example in which the designer uses color to indicate the function classes:

The primary design difference between these two column charts is using three colors to indicate the three function classes. This little change makes it much easier to recognize the ending of one class and the start of the other.

Color doesn't have to be limited to column areas. The following example extends the colors to the axis labels:

Again, just a smallest of changes but it makes a big difference.

***

It bugs me a lot that the long axis labels are printed in a slanted way, forcing every serious reader to read with slanted heads.

Slanting it the other way doesn't help:

Vertical labels are best read...

These vertical labels are best read while doing side planks.

***

I'm surprised the horizontal alignment is rather rare. Here's one:

 

Comments

[Andrew Gelman]

Kaiser:

I too have wondered why horizontal alignment is so rare---although we are seeing in now in coefficient plots in political science, and indeed you see it in the original graph B above.

I have a two-part theory as to why horizontal alignment is rare. First, vertical bar charts are standard, for whatever reason, and so they keep getting done (the qwerty effect). Second, we're trained to put the predictor on the x-axis and the outcome on the y-axis, and that's a useful convention; it makes it easier to quickly read a scatterplot or lineplot. Vertical alignment of the axis labels is what happens if you have a predictor on the x-axis whose values are character strings. I agree that horizontal alignment is the way to go. Communication is about tradeoffs, and in this case I think the benefits of readability outweigh the costs of abandoning the x-axis-as-predictor default. But I don't know that most people think about tradeoffs and choices when making graphs. I have the impression that, with graphs as with statistical analyses, most users think there's a single correct approach, and when they've done that, they'll stop.

[Kaiser]

Andrew: Agree on both points.

The qwerty effect is an interesting case study of tradeoffs - in the short term, it would be easier for each conforming reader to process the canonical chart form, and it would be more expensive for each dissenting reader to learn how to read the new chart form; in the long run, the time spent learning the new form will be rewarded but how does one convince someone to take a short-term hit for a long-run benefit?

[spilled graphics]

"These vertical labels are best read while doing side planks." haha well put. God is in the details. God is in the tweaks.