The good, the bad, and the really bad of data viz

When I see a really great chart or graph (data visualization or “data viz”) … or a really terrible one, I’ll take a screenshot of it.

I’m cool like that.

But here’s why I do it.

The really good ones get me excited, and the really bad ones make me so angry.

I know that not everyone is a data nerd like I am, and not everyone has the training, attention span, or interest to really determine what a visualization is telling them.

Most people just trust what they see to be true.

Last February, I saved the following image because it sent me into a fiery rage.

It’s not only confusing, but also irresponsible to not label the LARGEST part of this pie chart … which just so happens to be “Strongly disapprove.”

So while their messaging here might be true, that when categories are combined, more people at that time approved of Trump than not, more people actually chose “Strongly disapprove” than any other category.

And that was just buried in a whole bunch of words instead of intuitively labeled.

That’s my really bad example. Here’s another bad one.

When I saw the following bar chart a couple weeks ago with this headline, I was puzzled.

Why are we highlighting the 4th choice for who people are close to, which is barely a majority?

There are much clearer takeaways here (i.e., with spouses, partners, and non-family) to showcase.

If the focus of this report was on sibling relationships, then this is just not the right data visualization.

Neither of these examples make it clear, at a glance, what the data actually showed.

Those of us who are reading/scrolling quickly are at risk of interpreting the data incorrectly because of the visualization choices made here.

So when I talk about building capacity around data use … this is some of the stuff I’m talking about.

Of course, I want the people making posts with data for public consumption to do their jobs more effectively.

But I also want people to become better and more attuned CONSUMERS of data.

You don’t have to be a statistician, but under my watch, I want to make sure you see through the tricks that visualizations like these are playing.

But let’s end this on a good note because there is a lot of awesome data viz in the world too.

Thankfully, shortly after I saw the family relationships chart, I came across a carousel of images on Zohran Mamdani’s first 100 days as Mayor of NYC.

And they were *chef’s kiss* … SO GOOD.

Like this image on the left … it’s so simple, the graphics are relevant to the content, and there is no mistaking what is being visualized.

Or this series below: clear graphics, no complicated math to understand, and a serious impact as you scroll.

In addition to these awesome social initiatives for families, the Mamdani administration is also making sure that people can see and understand the changes being made in their city.

So here’s your reminder to be an astute and discerning data consumer!

And if you’re presenting your own data, think through how others might interpret what you’re sharing so you help — not harm — the perception of your work.

If you need help with this, let’s chat.