Maybe the numbers simply went up over time. Or down.

Other times, it’s more complicated.

Here’s how to explain multiple takeaway messages in presentations: with multiple slides, one per takeaway message.

Before: Everything Smushed on One Slide

Here’s what I typically see: lots of possible takeaway messages shoved into a single graph on a single slide.

The presenter says something like this:

“Next, let’s talk about gonorrhea diagnoses in our state. We’re looking at the number of diagnoses per 100,000 people. We’re also looking at age ranges. This is the person’s age when they were diagnosed with gonorrhea. We’ve got five years’ worth of data: from 2018 through 2022. Let’s look at a few key findings. Gonorrhea diagnoses were highest for people in their early twenties. In 2020, for example, there were 735 gonorrhea diagnoses per 100,000 people ages 20-24 in our state. Gonorrhea was lowest for ages 40-49. In 2022, for example, there were 88 diagnoses per 100,000 people ages 40-49 in our state. Here’s another pattern we found: Gonorrhea diagnoses generally went up from 2018 through 2020. For the three younger age groups, at least. And, gonorrhea diagnoses went down from 2021 to 2022 for all age groups.”

And while you’re talking through allllllll those numbers and age ranges and timeframes, the audience only sees this:

You see the problem, right??

The presenter is talking about one thing… but the audience is probably looking at something else.

That’s the very definition of Death by PowerPoint.

After: Describing One Takeaway Message at a Time with Multiple Slides

Instead, let’s use multiple slides!

We’re aiming for a single takeaway message per slide.

That way, what we say = what the audience sees.

There should be a perfect cohesion between sight and sound.

The presentation would look and sound like this:

“Next, let’s talk about gonorrhea diagnoses in our state. We’re looking at the number of diagnoses per 100,000 people.”

“We’re also looking at age ranges. This is the person’s age when they were diagnosed with gonorrhea.”

“We’ve got five years’ worth of data: from 2018 through 2022.”

“Here are the patterns at a glance. Next, let’s look at a few key findings.”

“Gonorrhea diagnoses were highest for people in their early twenties. In 2020, for example, there were 735 gonorrhea diagnoses per 100,000 people ages 20-24 in our state.”

“Gonorrhea was lowest for ages 40-49. In 2022, for example, there were 88 diagnoses per 100,000 people ages 40-49 in our state.”

“Here’s another pattern we found: Gonorrhea diagnoses generally went up from 2018 through 2020. For the three younger age groups, at least.”

“And, gonorrhea diagnoses went down from 2021 to 2022 for all age groups.”

Finally, you’d show the “full” graph again, pausing for questions and a discussion.

The Bottom Line: Use More Slides!!!

We’re not making the presentation longer. We’re speaking for the same amount of time as before.

We’re not rushing or slurring our words. We’re speaking at the same pace as before.

We’re not wasting paper or ink. These are the slides shown on screen during a presentation. If you want to print something, just print the “full” graph (which is slide 462 in the screenshot below).

We’re syncing our words and visuals.

We’re keeping our audience’s attention.

Because if they’re not even paying attention… How will they possibly understand, remember, and use the findings for decision making?!

The post When Graphs Have Multiple Takeaway Messages appeared first on Depict Data Studio.

There’s a misconception that accessibility takes all day, that’s it’s costly, or that it’s complicated. Those are all false.

Accessibility is woven into all my trainings, but since this is a topic I get asked about a lot, I decided to make a new talk that’s focused just on accessibility for dataviz.

In Spring 2021 I gave a talk at the Good Tech Fest conference about dataviz accessibility quick wins.

The talk was a “Choose Your Own Adventure” style where the audience chose what we discussed from a list of options. They chose:

• direct labels,
• lower the reading level, and
• lower the numeracy level.

You can watch the recording or read the highlights. Enjoy!

—–

Watch the Conversation

Here’s the main takeaway message: remove legends and directly label instead.

You probably know what a legend is, but direct labeling? What is that?

Let’s look at an example of a regular (inaccessible) graph.

Why Traditional Legends Don’t Work

When I saw this graph a few years ago, I actually liked most aspects of it.

I really liked the title in particular, and how it was phrased as a question, which gets the audience to engage. Two thumbs up to the title, “What happened to women in computer science?”

Legends Take Too Long to Read

But then I kept reading a little bit and I was like, “Wait a second… Time out.”

In full color I could mostly tell which section of the legend corresponded with which line. The turquoise lines were tricky because it’s hard figure out which is dark, which is medium, and which is lightest. Your eyes zig–zag back and forth trying to differentiate between the three. It’s really time-consuming.

Legends Don’t Work for Grayscale Printing

So it works in color, kind of, but what about grayscale printing?

Some people will view our graphs on-screen. Others will print them.

And if they’re printing the graphs, we should plan for grayscale printing. Colored ink is so expensive.

It doesn’t work at all.

Legends Don’t Work for People with Color Vision Deficiencies

What about color blindness?

If somebody has a color vision deficiency and can’t differentiate between red and green, the lines would all look yellow.

Traditional legends don’t work; they’re a thing of the past.

So what to do instead?

Directly Label the Graphs

We’re going to directly label our graphs.

What does that mean?

Direct labeling means you put the labels as close as physically possible to the data.

In this line chart, you’d just add the labels off to the side of the line.

Direct labels are:

• Faster for everyone to read (less eye zig-zagging)
• Grayscale-friendly
• Colorblind-friendly

A win-win-win!

Bonus points if you color-code the text to match the line it is labeling. (Red text for a red line, turquoise text for a turquoise line, and so on.)

How to Label Pie Charts

We’ve looked at line charts.

So, how do we label a pie chart?

Friendly reminder: Pie charts aren’t evil. They can be used as long as you follow the rule of two: you’re only allowed two slices in your pie. Maaaaybe three. The dark slice will be what you want the viewers to really look at, versus everything else in gray. Simple, right?

But we still need to directly label them, and it’s as easy as putting the labels as close as physically possible to their slices.

For example, if you have short labels, you can place the labels on top of the pie slices.

Now it’s speedier for people to read, it’s legible in grayscale, and it’s even legible for people with color vision deficiencies.

A question I get a lot is, “But if I have really long labels?” I know most of us aren’t comparing A to B.

If you have long labels, you can put your labels outside of the pie charts.

Bonus points again if you color-code the labels to the corresponding slices.

How to Label Donut Charts

Here’s another scenario for you with donuts. You’ve seen these, right? They’re just a pie chart with a hole punched in the middle.

They have the same rules as pie charts: two slices (max), with one dark slice versus everything else.

But, it’s really hard to fit any labels on top of donut segments. So how do you label these?

You have three options:

1. Outside of the donut segments
2. Inside the donut itself
3. Beside the donut

How to Label Bar Charts

Have you ever seen this, where Excel gives a legend that reads something like ‘Series1’?

This is confusing for viewers. To fix it, all you need to do is delete the legend.

How to Label Clustered Bar Charts

If your bars are long enough, you can place the labels on top of the bars, like this.

No need to label every single bar. Teach the viewers how to read the chart by labeling the top bars. Then, let them read the rest on their own.

During the Good Test Fest talk, an audience member asked how I added those labels.

You can:

• Add text boxes on top of the bars (beware: clunky and time-consuming)
• Use fancier automation techniques (e.g., concatenating the words and numbers together, a technique from this course)

How to Label Clustered Column Charts

I’m not a fan of putting the labels on the columns. The labels would need to be rotated vertically, which takes longer to read than horizontal labels.  

I typically use horizontal clustered bar charts to allow for horizontal labels, which are the fastest to read.

Download the eBook

Want to learn more about accessible data visualization?

In this ebook, you’ll learn 10 quick wins for designing accessible data visualizations. These small edits can have a big impact for our coworkers, board members, and funders who have color vision deficiencies, hearing loss, or learning disabilities–and for all of us who are pressed for time.

For your complimentary copy, use code: goodtechfest

The post Accessibility Quick Wins: Remove Legends and Directly Label appeared first on Depict Data Studio.

As part of the most-practical-keynote-you’ve-ever-heard emphasis, we included makeovers from the grantees’ real projects. In case it’s useful for my blog readers, I’m sharing one of those makeovers with you today.

Before: A Table

This table came from a grantee’s slidedeck. I’ve anonymized the county names where the grantee was operating, hid the program’s name, and changed the numbers of families, but hopefully you can get the gist of it anyway. This group displayed four months’ worth of data at a time because they held quarterly Board of Directors meetings. They opted to display the current month’s data (December) plus the prior three months (September, October, and November). We both agreed this was the “right” amount of historical context for their Board.

After: A Better Table

At the bare minimum, I would rearrange the table. I transposed the table–rotated the table so that the columns and rows are swapped–so that the time periods go horizontally from left to right. I always display ordinal data like time periods from left to right for consistency.

Then, I would declutter the table. Removing the background fills makes the data easier to read. There’s more contrast between the black numerals and the white background. Removing the background fills also allows me to overlay a heat map later on if I choose.

Finally, at the bare minimum, I would apply this grantee’s branding. They used Calibri and this exact shade of green.

After: A Line Chart

A second option is a line chart. Here’s Microsoft’s default:

Now, declutter that default graph! Remove the border:

Declutter the vertical axis and grid lines. In a moment, I’m going to add numeric labels to individual data points, so the axis doesn’t need to demarcate 0 and 20 and 40 and 60 and 80 and 100. And with exact values labeled, the grid lines no longer have a purpose, so they’ve got to go.

Add numeric labels through the center of each line–Edward Tufte’s graph table, my favorite chart type–as demonstrated in my tutorial.

Delete the legend and place the category labels directly beside the line:

Add a title. For live presentations, I opt for short titles that describe the graph’s contents but don’t give away the story. In many sense-making presentations, that’s the whole point of the meeting–to interpret the graphs together. My voice and talking points would describe the “so what?”–that County 2’s line went down while County 1’s line went up.

For handouts, I add subtitles. When I’m not physically present to discuss the graph, I type my talking points into the text box right below the title. There are always folks with schedule conflicts who can’t attend the presentation. Subtitles ensure that they know what you highlighted from each graph.

Match the font to the organization’s branding (Calibri, in this case).

Match the colors to the organization’s branding. Rather than using my navy and orange, I matched the organization’s style guide, which used these particular shades of green and orange.

Make sure your graph’s still legible in grayscale. Don’t wait until 4:59 pm for a 5 pm deadline! Print a draft a few days ahead of time. Compare the shades of gray. Can you still tell the lines apart?

Make sure the graph’s still legible for people with color vision deficiencies. Colorblindness is pretty common. It affects roughly one in ten people so you absolutely work with people who are colorblind and you should absolutely strive to make your graph accessible for those people. Head over to www.color-blindness.com and use their Color Blindness Simulator to preview what your graphs will look like for people who are red blind, blue blind, or green blind, among other scenarios. My orange-green graph turns into an orange-brown graph for people with red-green color blindness.

After: A Clustered Column Chart

A third option is a clustered column chart. This is the go-to chart for a lot of the people I work with, and I’d like to show you why it shouldn’t be your go-to anymore.

Obviously we’re not going to keep those default settings. I decluttered the graph, wrote a title and subtitle, and applied the grantee’s branding.

Make sure your graph’s legible when printed in grayscale. This is where the clustered column chart falls short. Clustered charts often need separate legends, and it’s much too difficult to distinguish the shades of gray apart from each other.

The separate legend doesn’t help anyone with color vision deficiencies, either. The viewers with red-green colorblindness would have to spend absurd amounts of their precious attention on distinguishing those muted oranges and browns apart from each other.

In theory, we could delete that separate legend and place the category labels directly on top of the columns. But then the text is sideways, which takes longer to read. And yes, in theory, we could transpose the graph (i.e., flip the graph from columns into rows). But then our ordinal data wouldn’t flow from left to right like the rest of our slidedeck.

Yes, direct labels hold up better in grayscale.

And yes, direct labels hold up better for people with color vision deficiencies.

Both legends and sideways text take longer to read, a drawback we can’t ignore… Yet another reason I despise clustered charts.

After: A Stacked Column Chart

A fourth option is a stacked column chart in which the County 1 and County 2 numbers are stacked on top of each other. Microsoft gives us this default:

In this edited version, I decluttered the graph, added a title and subtitle, and applied the organization’s fonts and colors for a hint of branding. I also added the totals on top of each column. Software packages call this a stacked column chart.

If our viewers care about proportions, we could convert those raw numbers into percentages. Software packages call this a 100% stacked column chart.

Deleting the legend and placing labels directly on top of the columns means that our graph would do fine when photocopied in grayscale. Notice the intentional white line between the green and orange counties, which helps to distinguish the shades of gray from one another.

The directly-labeled columns also hold up for people with color vision deficiencies.

Your Choice of Charts Depends on Your Message

“But Ann, which of these choices is correct?!” They’re all correct. Yes, all of them. Your choice of charts depends on your message.

The table puts viewers in the driver’s seat. Viewers have an opportunity to interpret data for themselves and come up with their own messages. I use tables for internal audiences, e.g., when you’re bringing data to your staff meeting in which the whole purpose of the meeting is to think about what the numbers mean.

The line chart shows whether lines are going up, going down, or holding steady. If you want to focus on County 2’s decline and County 1’s upswing, then this is the chart for you.

The clustered column chart directly compares the two columns. If you want to focus on the difference between the orange column and the green column, then place them beside each other in close physical proximity.

The stacked column chart focuses on part-to-whole patterns–how the orange segment and the green segment add up to a total. If you want to focus on combined numbers (289 families) vs. on break-outs (107 in County 1 and 182 in County 2), then the stacked bar chart is for you.

It all depends on your message. Sometimes you know the message beforehand. You might be trying to sway an audience to adopt a particular course of action and need to find data that support that course of action. Other times, you brainstorm several possible charts share those drafts with a colleague, and then choose your chart (and therefore your message).

Your Turn

I brainstormed four options for visualizing this dataset. Can you come up with additional ideas?

As usual, you can purchase the templates for the table, line chart, clustered column chart, and stacked bar chart.

Purchase the templates ($5)

The post How to Transform a Table of Data into a Chart: Four Charts with Four Different Stories appeared first on Depict Data Studio.

Most of these charts from my prior work were ineffective. Of course they were. Data visualization is a new skill for most researchers and evaluators.

One of my old charts was so horrific that I just had to share it with you:

What’s so bad? you ask. I see these all the time, you say. Can’t they just tilt their head to read the words, you argue.

What’s Wrong with my Previous Chart

Each of these mistakes–from the before I knew better period of my career, never to be revisited again–kept viewers from understanding the information:

• Generic and centered title (Should have a “so what?” and be left-justified.)
• No subtitle or annotations (So the viewer can’t skim the information.)
• Border (Should be removed.)
• Full grid lines (Should be lightened or removed altogether.)
• Legend (Should have direct labeling instead of the legend to avoid those back-and-forth eye movements.)
• Speaking of direct labeling… why did I label the axes and the bars? Overkill.
• Diagonal text (Should swap this vertical bar chart for a horizontal bar chart, leaving more room for the labels.)
• Default color scheme (Should use the client’s brand colors).
• An action color is there, kind of. Not really. It’s used incorrectly. Our brains are drawn to darker colors. Why was I drawing attention to the “possible items correct” section with the dark blue?
• Most importantly, why did I display the raw scores vs. the percentages? Can you imagine how much mental energy it must’ve taken my viewers to figure out what 2.9 vs. 2.6 vs. 6 means?

Before/After Makeovers

The first phase in learning about data visualization is usually critiquing charts. What, exactly, is wrong with your work or someone else’s? How can you learn from those mistakes?

The second phase is articulating specific ideas for how you’d make charts better.

The third phase is actually remaking your charts.

Some people stay in the first phase forever. My goal is to move more people into the third phase: actually improving your work.

So I’ll lead by example. Here’s how I’d re-do this chart today. I designed five different remakes.

Makeover #1: A Horizontal Clustered Bar Chart

First and foremost, I changed the chart type from a vertical column chart to a horizontal bar chart. Horizontal bar charts are great when the data labels are pretty long, like in this example. Your goal is to avoid diagonal text at all costs. It’s harder to read, so viewers get distracted, bored, or generally turned off and stop trying to decode your messy chart. I’m not a fan of vertical text either. (Not sure how to transform your vertical bar chart into a horizontal bar chart? I’ve got a tutorial.)

Next, I transformed the raw scores (“2.6”) into percentages (“43%”).

Finally, I addressed all the Low Hanging Fruit formatting issues. For example, I added a 10-word title and a 1-sentence caption; I deleted the border, grid lines, and tick marks; I placed the percentages on the bars so I could delete the axis; and I placed the years directly on the bars so I could delete the legend.

Why the black and white? Just for fun, I want to show you that you can still emphasize patterns without using bright, showy colors. In real life, would I match the action color to my client’s RGB codes? Absolutely.

What do you think of remake #1?

Makeover #2: A Clustered Stacked Bar Chart

I’ve experimented plenty with real-life clients to see whether they prefer regular bar charts or stacked bar charts. Here’s the typical response:

• Me: Here’s the first chart. [Regular bar chart.] What’s the message here?
• Client: Oh wow! Our participants are doing so well! 46%! That’s high! … Right? Or is that number low? Out of what? Out of 100%, right? Hmm…
• Me: You’re on the right track. Here’s the second chart. [Stacked bar chart.] What’s the message in this one?
• Client: Oh darn, we’ve got a long way to go. 46 out of 100%?! I need to speak with our program director about this. In fact, our whole team better see this. We need to figure out what we’re doing wrong before it’s too late!

A dozen conversations later, and the result is still the same: Nearly all my clients gain deeper insights about the findings through stacked bar charts instead of regular bar charts. What’s the response in your projects?

Makeover #3: A Small Multiples Bar Chart

The first two remakes are better than the original. That being said… clustered bar charts are my least favorite chart in the history of the world. They’re so cluttered. And worse, the comparisons are lost. With so many bars smushed together, it’s nearly impossible to see at-a-glance patterns between the two series of data.

In this example, I created a side-by-side bar chart so viewers could more easily see 1) the 2009 pattern on its own, 2) the 2010 pattern on its own, and 3) the difference between the two. Still in the bar chart family, but different patterns pop out, don’t you think?

I also used the action color (dark gray) to emphasize the Social and Ethical scores, and I added an annotation (the call-out box on the chart) to make my viewer’s comprehension even easier.

Want to make your own side-by-side bar chart? I’ve got a tutorial.

Makeover #4: A Dot Plot

Dot plots are often the superior chart. I use them to compare two points in time (like this example); two distinct groups (Program A and Program B); or, when I triangulate data, two distinct viewpoints (students’ perspectives vs. teachers’ perspectives).

But they’re not always superior. This dot plot doesn’t work. It’s too cluttered, isn’t it? Its more confusing than helpful. It’s because students improved on some areas, declined on other areas, and didn’t change at all on other areas.

The annotation isn’t helping, either; instead of adding clarity, it adds clutter.

For the rare viewer who’s willing to spend 60+ seconds interpreting the chart, it’s great because it shows more nuanced patterns than the other charts. But my guess is that the majority of viewers will lose interest after a few seconds because they can’t immediately grasp what it means.

Makeover #5: A Slope Graph

A slope chart is basically a line chart for two or three points in time.

This chart type is also effective at showing rankings (i.e., it’s easy to see how the skills are ordered on the left-hand side). 

For me, this chart is the winner. It’s easiest to understand at-a-glance. It ranks each of the skills areas. It shows differences over time. There’s a lot of cool stuff going on for the viewer to explore. It’s supplemented with a non-intimidating title, subtitle, and annotation. Yet, the information takes up very little ink and space.

Your Turn

Which remake would suit your audience best?

The post Avoiding Diagonal Text in Your Charts appeared first on Depict Data Studio.