But it doesn’t have to be!!!

Let’s stop hiding important qualitative concepts inside Text Walls that no one reads.

Step 0: Take Pride in Your Report’s Formatting

We’ll use landscape so that it’s easier to see on a landscape computer screen.

We’ll add the organization’s Theme Colors and Theme Fonts so that the document looks professional and enhances branding.

We’ll add a Text Hierarchy. Instead of a single Heading 1, we need to add several Heading 2s. We also need to enlarge all the font sizes. No more puny size 12 or 14 for headings!!

In this example:

• the Heading 1 is size 24 bold in a brand color,
• the Heading 2s are size 16 bold in a brand color, and
• the body font is size 11 in dark gray. (Not black, which causes eye strain and makes people think of funerals, at least according to my graphic design friends.)

Step 1: Choose Your Diagram

I like to scroll through SmartArt for ideas.

You can also browse Diagrammer, which is SmartArt on steroids.

Here are the most common diagrams I’ve used to visualize qualitative concepts in research and evaluation projects:

Processes

Processes are for linear, step-by-step concepts. There’s a defined start and end.

Examples:

• A lot of my own training programs, where I teach how-to instructions for dataviz.
• Logic models.
• Research methods (e.g., we recruited participants, and then they did this, and then they did this).

Cycles

Cycles are for processes that loop around and around until infinity.

Example:

• The program evaluation lifecycle, in which you plan for the evaluation, collect the data, analyze the results, use the data to inform decisions… and then start the process all over again.

Concentric Circles

Concentric circles are for spreading concepts and for inner, middle, and outer layers.

Example:

• An agency made a plan to improve diversity, equity, and inclusion. They identified three layers of changes needed: at the individual level, at the departmental level, and at the agency level.

Components

Components are for pieces of the whole—when you want to show that all these random things aren’t so random; they’re connected. They’re just not connected as a linear process or as a cycle.

Examples:

• In my master’s thesis, I researched how nonprofit organizations were using data to have a bigger impact on the community. In the literature review, I identified ~10 specific examples of data use, which were all related to the broader theme of data use.
• The U.S. Surgeon General’s Framework for Workplace Mental Health & Wellness (scroll down a bit here).

Pyramids

Pyramids, or ladders, are for concepts that build upon one another. The base layer is the foundation, the middle layer builds upon it, and you’re aiming for the pinnacle at the very top.

Example:

• In my Report Redesign classes, I organized the techniques into a pyramid. Participants learn the foundational skills, then the slightly narrower skills, then the nitty-gritty details that pull everything together.

Matrices

Matrices are fancy tables or plots.

Examples:

• This blog post about four types of dashboards.
• Lots of diagrams from McKinsey.

Venn Diagrams

Venn Diagrams are for interwoven, overlapping components.

Example:

• A project involving several groups of people, who all come together to advocate for their issue.

Honeycomb

Honeycombs, meh. I don’t love these. They’re overused, along with the gears. If you’re not sure what else to use, this is still better than a Text Wall.

Step 2: Introduce Your Diagram

Show the fully-colored diagram.

Don’t cram too much text on the diagram itself. In this example, I’m pretending we’re describing three steps, which repeat over and over.

Add a paragraph or two to explain the diagram at a high level.

Make sure there’s plenty of color contrast by using bold white or bold black text against your brand colors. Use this color contrast checker to figure out which font color to use.

Step 3: Repeat Your Diagram

Here’s the important part: Repeat your diagram as you explain each segment in more detail.

Copy and paste the diagram.

Then, gray everything out, and just highlight the segment you’re explaining in a dark brand color.

For bonus points, you can color-code the Heading 2s to match the diagram.

Make sure your colors are consistent with what you introduced earlier!!! You wouldn’t want Step 1 to be purple, and then blue, and then green.

I usually delete the words from the diagram that I’m not currently explaining. For example, when explaining Step 1, I delete the words Step 2 and Step 3 from the diagram. I don’t want any issues with color contrast; the white font wouldn’t be legible against the light gray diagram, so I simply delete it.

Make sure there’s plenty of white space between sections. I use at least 0.5 inches of white space (e.g., between the diagram and its paragraph, between the paragraphs).

For bonus points, break up the paragraphs into points and bold a few key words. Long paragraphs are dated. Readers expect short, skimmable paragraphs these days.

Check out the paragraphs in this blog post, for example. They’re 1-4 sentences long. There are lots of headings. There’s bolding to increase skimmability.

This blog post is also written at a 7th grade reading level.

Peek at the document with the gridlines on. Make sure the diagrams are aligned with each other.

The Final Product: Repeating Diagrams

I don’t care that it takes up two pages instead of one.

Two great pages will beat one lousy page any day of the week.

Yes, your boss might give you a made-up page limit. “Make sure everything fits on a page!” Those limits were created because bosses got tired of Text Walls. And, because we used to print a lot.

Nowadays, people don’t print as much. I think the pandemic was a major turning point. With everyone working remotely, nobody had access to the office printer anymore. Any who wants to pay to print at home??

I’ve never, ever heard complaints about two accessible pages vs. one inaccessible page. The word count is the same. (Well, I added some headings.) But the information is richer because we’ve added a diagram and then explained it piecemeal.

Adapt as Needed

Use can use any diagram you need—a cycle, linear process, pyramid, or concentric circles.

You can do this in Word.

You can do this in PowerPoint.

You can do this in Canva.

You can do this in Publisher.

I’m software-agnostic. I don’t care which software program you use. When formatted well, you’ll get the same high-quality end result regardless of which program you’re using.

In this example:

• The diagram was wide, so when I introduced it, it needed the full width of the page.
• When I repeated the diagram, none of the words (“Phase 1”) fit, so I deleted them.

Adapt as needed!!!

Download My Word Document

Bonus!

Want to see how I arranged everything inside of Word?

You can download the document here: https://depictdatastudio.gumroad.com/l/UseRepeatingDiagrams

My friend Jon Schwabish from PolicyViz asked me to partner on his One Chart at a Time project, in which we’re helping you get better-acquainted with common and not-so-common chart types.

I created a tutorial on using colored phrases to visualize qualitative data. In this tutorial, you’ll learn:

• The first time I ever used colored phrases to visualize qualitative data;
• My favorite examples of colored phrases; and
• Practical tips for using colored phrases in your project.

Watch the Tutorial

The First Time I Used Colored Phrases in Data Visualization

In the first part of the video, you’ll see the first time I used colored phrases.

I was coding open-ended survey data during my Master’s thesis, and needed to visualize different themes that I was finding in the data.

Inside good ol’ Word, I added colored rectangles around key phrases.

The outcome wasn’t perfect, but it was better than regular text.

My Favorite Examples of Colored Phrases

In the second part of the video, you’ll see my favorite examples of colored phrases:

• ‘Stronger Together’ and ‘I Am Your Voice’—How the Nominees’ Convention Speeches Compare by the New York Times
• Did Michael Brown Charge? Eyewitnesses Paint a Muddled Picture by the Washington Post

Practical Tips for Using Colored Phrases

In the final section of the video, you’ll learn practical tips for using colored phrases to visualize qualitative data.

We’ll go through seven options:

1. Regular text
2. Bold
3. Italic
4. Underline
5. Color
6. Outline
7. Fill

You’ll learn the pros and cons of each approach, and see why I suggest using bold, colored, or filled text instead of the other options.

Your Turn

Let me know when you’ve applied colored phrases to your own project!

—–

I have been an enthusiastic consumer of data visualization workshops, blogs, books, and tips over the years, and I am always trying to take my data viz skills to the next level. I have been following Ann’s blog for several years now, and I love the way she presents material in a user-friendly way that makes you realize that good data viz is achievable.  

Last year I enrolled in Ann’s Great Graphs course to further my skill set and I’m really glad I did. I love not only her way of showing how to create great visuals in software we already have (Excel and PowerPoint), but she also has a great eye for design and presentation that sets her approach to visualization apart from other data viz gurus. 

What I learned in Ann’s course has been invaluable in so many ways. Not only have my skills improved, but I now approach every data viz project with the goal of telling a story by intentionally using data visualization design concepts to make the story clear. By taking the time and putting in the effort, my skills have improved and so has my work.  

Here are three ways my work has improved since taking Ann’s course. 

Improvement #1: Presenting Qualitative Data 

One of my biggest struggles as an evaluator is how to present qualitative data in an engaging and interesting way.  

I want my clients to take away key points from the qualitative data I collect and report. 

What My Qualitative Data Visualization Used to Look Like 

Unfortunately, I don’t think I used to achieve this goal.  

Here is an example of how I used to present qualitative data:

Yes, that’s right…page after page of text and quotes.  

Nothing there to make readers remember what they are reading or highlight the important points.  

I can only imagine my clients’ eyes glazing over as they waded through all this text to try and figure out the point I was trying to make.  

Even worse, I could imagine them skipping these pages altogether. 

What My Qualitative Data Visualization Looks Like Now 

Ann’s course gave me inspiration to change my approach in her lesson about presenting qualitative data, specifically, using color-coding to highlight key phrases and help readers home in on the most important pieces of information.  

My first attempt looked better and was more user-friendly, but my boss helped me step it up even more by inserting icons to help readers remember the key take-aways (by the way, Ann also encourages the use of icons).  

Below are the two newer versions, and I now use the icon format in many of my reports.  

This approach is so much better for my clients, as the icons represent and help them remember the main themes, and color-coding portions of the quotes draws attention to the phrases that illustrate each theme. Now my clients don’t have to struggle to pull out the main ideas.  

And, speaking of icons, see that icon at the top right corner of the page? Every chapter in this (very long) report had its own icon that was first used on the “Contents” page, and then repeated on every page of the chapter as a quick reference for my readers. Each chapter was a different color as well, which also aided in helping readers orient themselves in the report. 

Improvement #2: Presenting Pre/Post Data Using Dot Plots 

I do a LOT of pre/post data collection in my work.  

By far, one of my favorite ways to present pre/post data is with a dot plot.  

I’ve been making them for years, but I’ve always been a little disappointed in the results.  

For example, when some variable increase and others decrease, I never felt like there was a clean way to set them apart from each other.  

Yes, I tried used arrows and different colors, but I never quite liked the final result.  

Great Graphs changed that! Ann showed me how to create dot plots that are modern and that beautifully highlight when some variables increase, some decrease, and some don’t change at all.  

By color-coding the different groups, using arrows, adding some white space, and adding icons and descriptors, the graph is much clearer. 

What My Dot Plots Used to Look Like 

Here is one of my old plots. It’s good. But it could be better. 

What My Dot Plots Look Like Now 

Here is how I make dot plots now…so much more informative and easier to interpret!

Improvement #3: Making Report Covers 

I put a lot of work into my reports, so I want my clients to read them.  

One way to get them interested is by making report covers that are interesting and engaging.  

My old report covers failed miserably.  

Did they provide the necessary information? Yes…title, name of program, date, author, etc.  

But they were BORING.  

In Ann’s course I learned how to make better report covers by using full-bleed photos.  

Here’s what my old report covers looked like:  

What My Report Covers Look Like Now 

And here are some examples of what my report covers look like now – colorful, interesting, and engaging: 

The Results 

These are just three examples of how my work has been elevated by applying Ann’s tips, tricks, and ideas.  

Having these new skills in my arsenal makes me look forward to each new data viz project.  

I love creating visually appealing reports, and thanks to Ann’s course, I have the skills and inspiration I need! 

The vast majority of data visualization resources focus on quantitative data.

In this article, let’s look at some of your options for qualitative data visualization, like word clouds, photographs, icons, diagrams, and timelines.

What *Not* To Do

I used to conduct qualitative-heavy research projects pretty much all day every day.

Key informant interviews, bellwether interviews, document review, focus groups, you name it…

So I know from personal experience that 99.9 percent of qualitative reports look like this:

Which is fine, unless you want someone to, ya know, actually read your report.

Here are several ideas for presenting qualitative data for your organization’s reports, presentations, handouts, infographics, and more.

Word Clouds

Data visualization novices love to love word clouds, while data visualization experts love to hate word clouds.

Word Clouds Are Okay for Visualizing One-Word Descriptions, But Not for Visualizing Allllllll Your Qualitative Data

Here’s the most obvious strategy for visualizing text-based data: the word cloud, also known as a tag cloud. Frequent words or phrases are shown in larger, bolder font. Less-frequent words or phrases are shown in a smaller font.

I’m not advocating for run-of-the-mill word clouds where you simply dump your interview transcripts into a word cloud and hope for the best. Simply looking at how often a word or phrase appears in your dataset is not a sufficient way to analyze your data!

Instead, I’m talking about intentional word clouds – when you’ve considered other options, and have purposefully chosen a word cloud as your visualization of choice.

Here’s one example. When I conduct interviews as part of a qualitative research project, I often included a question in the interview protocol such as, “What one word you use to describe _____?” For example, “Mr. Policymaker, what one word would you use to describe public opinion towards poverty in the United States?” or “Mrs. Principal, what one word would you use to describe this new initiative’s effect, if any, on your charter school?”

The rest of the interview is obviously focused on loooooooong responses, and fuller descriptions, and plenty of examples. But for one or two moments during the conversation, it’s nice to pause and throw in a semi-qualitative short question.

Then, you’d visualize those one-word responses in a word cloud, like this one from Students First:

Here’s another example from the Pew Research Center, in which they visualized one-word descriptions of Barack Obama with packed bubbles. Packed bubbles are similar to word clouds. In word clouds, word frequency is depicted through font size. In packed bubbles, word frequency is depicted through bubble size.

Word Clouds Are Okay for Visualizing Before/After Comparisons

Word clouds are also great for before/after comparisons, like these tweets describing breakups. Does your study involve pre/post tests with a few open-ended questions? Did you interview participants at multiple intervals during the study? You could adapt this technique for nearly any time series design.

What our breakups look like on Twitter http://t.co/yRZ8fBgRxx Before and after: pic.twitter.com/XQg7JKMar5 — Motherboard (@motherboard) September 23, 2014

The Best Software Tools for Creating Word Clouds

There are tons of free and low-cost websites for generating word clouds. The Poll Everywhere blog summarized nine of their favorite tools. My personal favorites are Wordle and Tagxedo.

Wordle

Wordle is a free website for making word clouds in an oval or rectangular shape. It’s pretty easy to use. You simply copy all of your text (e.g., from your interview transcripts) and paste the text into Wordle. Then, you can customize the word cloud’s fonts, colors, and general shape to your liking. I suggest matching the fonts and colors to your project’s existing branding guidelines so that your word cloud matches the rest of your report or slideshow.

Tagxedo

Tagxedo is a free website for making word clouds in fun shapes, like an apple, heart, or outline of the United States. You can pull in qualitative data directly from a website or Twitter feed (you don’t have to copy and paste, so you save a step). And, with Tagxedo, you could even have your word cloud printed on a t-shirt, tote bag, or coffee mug. I see Tagxedo as more of a decorative option than a researchy option.

Showcase Open-Ended Survey Data Beside Closed-Ended Data

What’s better than quantitative data? Or better than qualitative data? Quantitative and qualitative data combined!

Use this technique when your survey has both closed-ended and open-ended questions. Tie the responses together in one chart to add much-needed context. In this example, a survey asked nonprofits to describe what it was like to work with an outside consultant. Then, the survey asked them why their experiences were good or bad. Who cares if 33 percent of nonprofits had excellent experiences but we don’t have examples that describe why?

Rather than simply listing out the open-ended responses in your appendix, showcase them beside your stacked column chart.

Read more about this technique in this post written by Johanna Morariu and I.

Include Photos Beside Participants’ Responses

For your non-anonymous reporting, how about inserting photos of the interviewees next to their ideas?

Strategies for use of #interpreter services in hospitals http://t.co/EpvPyvNcMZ, from @abtassociates @NHeLP_org &more pic.twitter.com/O8V05h3DoM
— Sabriya Rice (@sabriyarice) September 2, 2014

Here’s an example from the New York Times, where they added photographs beside interviewees’ responses:

Include Icons Beside Descriptions and Responses

Icons are so easy to use that there’s really no excuse for not using them to break up long sections of text.

Islamic State is one of the world’s richest terror groups. Here’s why: http://t.co/C0DsXqHkfD pic.twitter.com/ikNThUwNhb — Wall Street Journal (@WSJ) August 28, 2014

Another great use of icons comes from the Washington Post’s story about how animals are faring after the BP oil spill. Isn’t it remarkable how icons and bold subtitles help to break up a chunk of text?

The Center on Budget and Policy Priorities used icons to visualize how working-family tax credits can help at every stage of life:

And the Center on Budget and Policy Priorities used icons again to help their readers understand how low-wage workers face a cut in tax credits:

Create Diagrams to Explain Complex Concepts and Processes

You can create diagrams for many aspects of your project, like this diagram explaining what type of protective gear is keeping doctors safe from Ebola:

Breaking down the protective gear helping keep doctors safe from Ebola http://t.co/bQUDqFoD1X pic.twitter.com/tDjkiSLFvs
— Washington Post (@washingtonpost) September 15, 2014

Graphic Timelines

Regular text-based timelines + diagrams, photos, and other images = graphic timelines. Timelining is especially valuable in situations in which you need to track how a program, initiative, or campaign unfolds over time.

Interactive Timelines

This timeline from the New York Times includes an assortment of qualitative and quantitative information: dates, milestones, # of rockets, and # of casualties. My screenshot doesn’t do it justice; head over to the New York Times website to explore the interactive version.

Static Timelines

Maybe you’re not the New York Times, I get it. Anyone can design a static timeline from the comfort of their PowerPoint slide.

A Graphic Timeline of the 21st Century Conservation Service Corps http://t.co/GmpTR3qfqv #21CSC #conservation pic.twitter.com/nGydFtkXkC — The Corps Network (@TheCorpsNetwork) August 12, 2014

Color-Coded Phrases

One of my all-time favorite examples of qualitative data visualization comes from the New York Times’ election coverage in 2016.

They compared and contrasted speeches from Donald Trump and Hillary Clinton.

First, the New York Times team presented miniature thumbnail images of each nominee’s convention speech. You can even click on the thumbnail images and they’ll expand so that you can read the transcript. I love how the thumbnails provide a birds-eye-view analysis of the qualitative themes from each speech simply by color-coding certain phrases.

Directly underneath the thumbnails, the New York Times team pulled out a few sample quotes so that readers can get a sense of what was said.

Want to try this technique yourself? Color-coded text is deceptively simple. I suggest highlighting the entire phrase in a light color, as the New York Times has done in this example. Make sure you don’t just color-coded the font itself. It’s challenging for our eyes to read green letters, but it’s pretty easy for our eyes to read black letters against light green background shading.

Comment and Share Your Ideas

How are you visualizing qualitative data interviews, focus groups, surveys, document reviews, and other qualitative data sources?

Please link to your favorite resources below.

Let’s give virtual high fives to the creators and celebrate a job well done.