UK University Funding Sources_before-after

Disclaimer: By re-creating this visualization and sharing it publicly, I take no sides in the ongoing debate concerning funding for universities in the UK. I have taken the time to conduct some research in order to better understand the current situation and the arguments on both sides (brief summary provided below), but I cannot begin to claim that I fully understand what is going on. I simply engaged in an exercise in data visualization by taking an existing chart, breaking it down and making sense of it, and then crafting a new chart to better communicate the message that Phil Baty and Universities UK were trying to make.

I originally encountered the data visualization on the left via Phil Baty‘s Twitter post. He had cut and pasted a donut chart that Universities UK had created for their online publication, University Funding Explained (see also their Prezi, which has a slightly different and slightly better chart). My initial reaction was one that is all too common among those who work in the data visualization field: No More Pie Charts (or in this case, a donut chart, which is even worse)! If you’re unfamiliar, or even surprised, by the negativity directed toward pie charts, I’ll share some of the reasoning below.

When I looked more closely at the pie chart, I realized that I had no idea what the graph was trying to communicate. I reread Mr. Baty’s tweet multiple times to try and gain a grasp on the argument at hand, but 140 characters does little to inform someone completely ignorant of an ongoing debate. In fact, I emailed Mr. Baty to ask some clarifying questions, and only after hitting “Send” did I realize that I completely misunderstood his stance. So back to square one…

Essentially, there has been a lot of recent criticism directed at UK universities; one of the major voices is that of Lord Andrew Adonis. He has questioned rising tuition costs and the increasing debt that university graduates possess upon leaving school (sounds pretty familiar to me, as an American-college-graduate); he has also targeted the universities’ vice-chancellors, who are raking in more-than-decent salaries. In response, Universities UK has been trying to provide some education as to funding sources for the universities, pointing out that because the government provides only 26% of university income, the institutions have to look elsewhere, i.e., rising tuition costs, to close the funding gap. Furthermore, the argument goes, reduced government funding has privatized and marketized education, putting the vice-chancellors in a position analogous to that of senior-level business executives. If you want to learn a bit more, read this post by the Times Higher Education.

Now, let’s look at my decision-making process related to re-creating the visualization:

Here’s the original:

UK University Funding Sources_before
Things that work (to some degree):

1. The use of the two different shades of green ties the two government categories together.

2. The boxed caption highlights the key takeaway so that the reader doesn’t have to guess what that might be. However, the small font makes it hard to read.

3. The percentage labels allow the reader to quickly interpret the size of each “slice” without the need for guesswork. The labels are the same color as the slices, and they are in close proximity to the slices.

4. The donut chart allows “£33.2 billion” to appear in the center, allowing the absolute value to accompany the percentages. With that said, it does take a small leap of logic to understand that “£33.2 billion” represents the total income.

Things that don’t work:

1. The title gives us no idea what we’re looking at. Total income for what?

2. The rest of the color scheme does not make sense. I think the creator is trying to limit the number of colors used, while still distinguishing the different categories from one another. However, The two purple categories and the two orange categories don’t have anything to do with one another.

3. There doesn’t seem to be any intention behind dark versus light colors. In fact, light green for “Research — UK government” makes us gloss over it in favor of dark purple and dark orange. And, the light green quantity (16%) is actually larger than the dark green quantity (10%).

4. Despite the presence of a key, it is very annoying to repeatedly scan up and down to look at the donut, then at the legend in order to identify the six categories.

5. All of the percentage labels are the same size, except for the 1% label. Yes, that particular slice is very small, but the lack of consistency in label size doesn’t work.

Here’s what I created:

UK University Funding Sources_after.png

Decisions and Rationale:

I decided to use Adobe Illustrator CS5 to create my visualization. I like the ease with which I can enter data to create a chart, and then alter and add attributes as needed. There are some drawbacks to using Illustrator, but in this case, it served my purposes quite well.

1. I created a title that communicates what we want the reader to understand and look for as she views the visualization. If there is a story that we want to tell with our data, then we don’t want to leave it up to the reader to try to figure out the story.

2. I used a horizontal bar chart. The funding source categories precede each bar, taking advantage of the principle of proximity, and the relative sizes of the bars allow for easier interpretation of magnitude. (Looking back at the donut chart: without the percentage labels, 8% and 10% look very similar.)

3. I used color strategically to link related concepts. I used dark blue to link the title, and the two government categories (both labels and bars). I used gray for everything else. Furthermore, the contrast between the two colors focuses the reader’s attention on the government bars.

4. I put the percentage labels inside the ends of the bars, communicating the percentage represented by each bar.

5. I made the font size for the government bar labels (both descriptive and percentage) slightly larger than that for the other labels, an application of the principle of size to highlight important information.

6. I added a few other pieces of information, again taking color and size into account and using smaller font to not distract the reader from the main message. The black “Funding Sources” label tells us what the descriptive labels represent. Below the bars, I added the “Total income” for those who might wonder how much money we’re actually talking about (I, for one, hate dealing purely in percentages and want to know absolute values). I also added the source of the data.

What did I miss? What would you have done differently?