Labels, Titles and Endorphins, Oh My!

I have been an avowed hater of running since forever. Not wanting to run is just about the only thing in my entire life that I have ever been certain of.

That is, until a couple of months ago, when my associate and book researcher, Ann Cutrell, came to town for a few days.

Ann started running some years ago and now is an amazing triathlete performing at the top of her game. She is so good, in fact, that she recently placed 13th at the 2012 USA Triathlon Age Group National Championships in Burlington, Vermont. Here is a picture of Ann looking strong at the event:

Ann running

As I listened to Ann describe how much fun she is having competing in triathlons, I was (1) surprised at how much I loved hearing her stories, and (2) flabbergasted that I caught myself thinking about giving running a chance.

So now, much to my complete and utter disbelief, I am here to tell you that I decided…what the hell? — why not see if I could get myself in good enough shape to run in a 5K race, which is “only” 3.1 miles? (“Only” is such a deceiving word, isn’t it?)

I found a free app on my iPhone called RunKeeper, and set it up to track my progress on a two-mile loop around my neighborhood. It’s a very cool app with great functionality like GPS tracking and time-keeping. And if you use it with ear buds, you can listen to both music while you are running and to a very nice woman from the app who coaches you along: “Start running now…your average pace is pathetic…” (I made up the “pathetic” part, but you get the idea).

RunKeeper also provides lots of feedback in the form of data and data visualizations; however, those need a bit of improvement.

Here is one graph I pulled up just today to check my progress.

progress graph

When I looked at this graph, my eyes went immediately to the biggest bar, the one marked 27 August — and I was psyched! I had run six miles on August 27th. I didn’t actually remember doing that, but the graph showed that I had, so it must be true. And then I stopped to think about it. Had I actually run more than six miles in that one day? If so, then everything I’ve ever heard about endorphins must be true.

Then the cold, hard truth washed over me as I hovered over the bar and saw the following details:

week total
That impressive number — six miles — was for the entire week of August 27, NOT for the single day August 27. My moment of glory was dashed in one fell swoop.

I had fallen victim to something I warn my clients about all the time! I advise them on a regular basis to make certain their labels are crystal clear, in order that viewers of their reports will understand EXACTLY what they are seeing. Otherwise, the very real risk is that poorly labeled graphs will result in an incorrect interpretation of data — and worse, incorrect decisions (or, in my case, demoralization) based on that misinterpretation.

RunKeeper could improve the clarity of this graph by simply improving the labels as I have done in the following graph of the same data:

total miles per week

So what are the general guidelines that you should keep in mind when you are labeling your graphs of healthcare data? Here are my top five:

  1. Write. Simply write or type a few different titles and labels on a piece of paper and think about them.
  1. Correct. Make sure that the title and labels of your graph are consistent with the data being displayed. For example, don’t entitle your graph “Mean length of stay for medical patients” and then label an axis “Median length of stay.” Be certain that you know what values are being displayed and label them correctly throughout.
  1. Cut. Eliminate unnecessary words; make every word count.
  1. Test. Show your graph, with the titles and labels you think will work, to someone who has never seen the information, and ask, “What does this graph tell you? What do you see?”
  1. Edit. Just as you edit anything that you’ve written, you need to edit your labels and titles. Your first attempt can almost always be improved upon.

The power of data visualization is that it permits, even encourages, the viewer of a graph or chart to see and understand information — at a glance. Crucial to this effectiveness are unambiguous labels that inform and help viewers know and understand what they are seeing.

These guidelines are simple, and they work. In fact, now that I have applied them to my running graph, I am certain of something else apart from my feelings about running: Ann Cutrell needn’t worry that I will be claiming her triathlon crown anytime soon.

Share
This entry was posted in Best Practices, Data Analysis, Design Basics, Graphs, Newsletters, Tables. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *


6 + = nine