Data Visualization

Chart Guide: Dot Plots [graphic novel blog post]

December 3, 2019 Chris Lysy

A bar chart and a scatter plot had a little baby. In this short graphic novel guide I’ll walk you through dot plots, why they exist, and how to use them.

So just what are dot plots?
And how do you make and use them?

If you've never heard of dot plots it's because Excel doesn't give you that option when you click "insert chart."
But 3D Line Charts and 3D Pie Charts are options. Avoid the urge to make any chart 3D.

So dot plots are a kind of mix between bar charts and scatter plots.

Bar charts have a way of taking up a lot of visual space.
Dots are much less in your face.

They hep narrow your focus. Instead of looking at blocks your eyes focus on the dots.

For a bar to be accurate, you need to show the full length.
Whereas dot plots work based on relative position. Meaning you can zoom in without distorting the visualization.

You also get the chance to pack more data in the same visual space. Meaning you can compare across and within with the same chart!
Great for demographics like race, gender, ethnicity, income bracket, etc.

Dot plots are a standard option in Tableau. But they're easy enough to create in Excel if that's your preference.
Just create a sequential numbering column in your spreadsheet. Then just choose scatter plot with the data and the new sequential number column.

Bonus tip. Dot plots are also really nice ways to display pre-post data. You can even connect the dots to create little dumbbells. — *All typos/misspellings are intentional as a way to show humanity, not an example of self-editing laziness.

Want more step by step Excel instructions?

Stephanie has a couple of good guides:

Confession Time: I’m not a reader [graphic novel blog post]

December 2, 2019 Chris Lysy

In today’s graphic novel blog post I talk about the underlying challenge that drives my work.

I have a confession.

I am not a big reader.
Don't get me wrong, I like the idea of reading, and enjoy being read to.
But when I read to myself one of two things happens...

The book is really interesting and causes my mind to go off on tangents.
The book is really dull or repetitive and my mind just shuts off.

Now graphic novels on the other hand. I can't get enough of them. And the industry is booming.
Used to be mostly super hero comics.
But more and more we see history, memoirs, biography, adaptations, and literature.

Unfortunately for people like me, the world of research and evaluation tends to be one big block of text.

Except for some notable areas.
Data visualization, aka the most academically acceptable form of illustration.
Social media and digital dissemination.
Presentation design.
Modern report design.
eLearning design.

I spend a lot of time trying to find my place in the contemporary professional world.
Am I a data analyst?
Am I a designer?
Am I a digital publisher?
Am I a developer?
Am I a teacher?
Am I a cartoonist?
But i think the truth might be far more simple...

I think I'm just frustrated with the slow evolution of the academic world towards a more visual interface.

Because hidden in those big blocks of text are important ideas and evidence critical to making our world a better place.

And if we don't adapt and evolve, we create space for conmen to hijack the public trust.
[like Trump]

Now I know cartoons and graphic novels won't save us. But this needs to be better.

But I'm curious. Does this resonate at all, or is it just me?

Love t hear your thoughts in the comments, on twitter (@clysy), or on LinkedIn (LinkedIn.com/in/clysy).

Chart Choosing Advice [Graphic Novel Blog Post]

November 26, 2019 Chris Lysy

If we chose clothing the way we chose charts, we would all look the same. How dull. My contention, there is no single right chart or wrong chart, that said, here is how I go about choosing.

So how do you choose the best chart for your data?

I tried to ask Google, but the search results were inconclusive. So here's some advice based on my experience.

You can hardly go wrong using one of the old reliables. Line Charts, Bar Charts, Scatter Plots.

Just because you "can" put everything in one chart, doesn't mean you should.

Splitting up a stacked or clustered bar into separate charts can make it easier to process and apply focus on what is most important.

Stacked bars can be useful when comparing the totals is the main point.

But in some instances, like ratings and scales, shifting the central axis can really increase the value of the chart. Like in a deviated bar.

Bars are reliable tools, but they're not the only game in town. All sorts of other shapes can represent the same data points just using position.

So why switch out a bar for a dot? One advantage, you get to pack more data on the same line. This can be great for breaking down by variables like race and gender.

There are a lot of people who hate pie charts. But there really isn't anything wrong with basic pies. A simple pie chart is fine.

Now when you add too many slices, or turn it 3D. That's a different story.

With too many slices you can't possibly compare. And 3D charts distort data.

With the help of a lighter color or shade you can add more data to a chart but leave the focus. This can give your charts added depth.

So how do you visualize descriptive statistics? Charts like box and whiskers can confuse non-stats people.
When possible, I just visualize all the data. Then emphasize the median with color.

Trying to figure out the best chart is like looking for the perfect cookie recipe. The "Best" is highly subjective.

Try some different options, ask for opinions, and go with your gut.

So, if I were to create a bunch of these types of lessons and put them together into a graphic novel guide, would you buy it?