For the last two years, Viz for Social Good has ended the year with a project looking at its volunteers. Last year, it was all about the projects and who submitted them, and this year, it was looking at the results from an org-wide survey.

As some of you know, I was job seeking for most of 2020, so when I got a new job in 2021, a few things stopped. One of those things was my participation in Viz for Social Good. Suddenly I was juggling a full-time job, a side hustle, and trying to find time to viz on the side of the side hustle. Tough stuff.

But! I retooled my side hustle’s business model to free up my time (maybe another post on that another time!) and when the 2021 VFSG survey project rolled around, I dove in. (Scroll to the end of this post for the full viz.)

I won’t explain all my thought processes here, but suffice to say that even if I’m doing a quick viz, I try to do three things:

1. Pleasing color

2. Tidy, “finished” format

3. Something interesting in a viz

This time around, the “something interesting” was heat maps.

When I started looking at the survey results, I realized that looking at the two metrics gender and age could be misleading. Taken separately, they could mischaracterize the “average” volunteer. Consider the following:

In this table, we can see 20 sample responses. If you chart ages and genders separately, you get:

Check it out! Most respondents were male, aged between 18-25. Right?

Not exactly.

When you put the two metrics together in a heat map, the interpretation changes pretty sharply:

The most populous group here is women between 36 and 50, very much not men between 18 and 25. In reality, the men were more evenly spread in age, which is why they came out tops in the total. Women, in this case, were concentrated in age.

So does that mean you should do everything as heat maps? No! But it should be a consideration when you’re presenting your results. Are you arguing that VFSG should advertise on Joe Rogan? Might be worth knowing your demographic is middle-aged women, not young men! But if you’re just debating whether to spend your ad budget on men or women, just the straight gender total will do.

In my case, I chose three heat maps:

1. Age and gender

2. Job title and experience

3. Hours vizzing at work vs hours vizzing for VFSG

These were the comparisons that worked for what I wanted to show, so that’s what I went with.

Plus, it worked for the design.

Check out the final viz on Tableau Public, or below.