I thought about writing a post describing this new series I will follow. Instead, I decided to use it as a curriculum for data visualization. Let’s start with the graph, which will open in a new window or tab. The focus of your eyes should immediately go to the red and green lines. Those are the important elements of the graph and it should be what you notice first.
Let’s move beyond that and talk about why they are the most noticeable. They are a brighter color compared to the rest of the items on the graph. The right-hand side is clear of other items and that space gives us a clear view of the most recent data observations of the two series.
After that, we may look to the horizontal or vertical axis — the horizontal axis because we are interested in the time of the observation or the vertical because we are interested in the scale of the data. In both cases, there are numbers without text. I have been a strong advocate for labeling each axis until recently when I noticed people using text within the graph to describe the axes. There are two advantages to this method: text is not rotated 90 degrees for the vertical axis and the text box doubles as the legend. In effect, we have replaced three text boxes (the horizontal axis label, the vertical axis label, and the legend) and replaced them with a single text box that may not have any more words than the composite of the original three.
In this particular case, we see what the horizontal and vertical axes are. We can also see a legend which shows the two data sets using the color of the line as part of the legend. This removes the graph type icon from the legend of most software and doubles the density within the new legend — a single line describes the series and defines which series it is.
Perhaps next our vision moves to the center of the graph which shows another text box with the most recent observation. I generally approve of this on all graphs since it shows how current the data is as well as the value of the most recent observation. It is not always easy to tell what each observation is especially as the line moves farther away from the scale which is generally on the left-hand side. This text box leaves no doubt on the value at the end point of the line. Notice that I also included the color of the line for each data observation. While this is not necessary for this graph, it provides reinforcement on the last observation for each data series to match it to the proper line.
By now, we have noticed the large paragraph of data near the left-side of the graph. I generally do not like detailed text on graphs. I especially do not like them when I am including a lecture — if I am truly speaking to the graph, there is no reason for additional text to distract from my comments. The counter to the lecture is the printed graph, which is what we are seeing now. For this type of presentation, it might be useful to include an explanatory paragraph to provide detail into how to read the graph and why it is meaningful. Since this graph will be updated at best quarterly, including background and detail will be helpful for any audience.
Finally, there is a text box in the lower-right which shows the source of the data. I am usually quite deficient at successfully including the data source. It is best to include it because you will have skeptics who will not believe your conclusions and observations until they look at the data series themselves.
Okay, that is it for the text boxes. They add a lot to the visual stimulation of the graph, but I think they have their purpose and they do not detract from the data. There is one item to discuss that is not seen — the gridlines. I did not speak of them because they do not stand out. Tableau defaults to a light gray color and I highly approve. Microsoft Excel defaults to black and it can make it unnecessarily difficult to read a graph. In this case, they are subtle and come into visual acuity only if needed.