Qualitative Data Visualization: The Gauge Diagram

guest post by Jennifer Lyons

When we debuted our Qualitative Chart Chooser, we promised to dive into detail on specific visualizations, so let’s kick it off by discussing how and when to use one of the most derided charts of all: the gauge diagram. I know, I know… I am sure you are tired of them by now after being stressed out by the New York Times’s jittering gauge showing the live presidential forecast. Most of the time I think gauge diagrams are useless. I am going to take a different spin on this controversial chart type by introducing two scenarios where I think they can be useful for qualitative data.

First, let’s start by covering why people hate these diagrams so much. They take up a lot of space on a dashboard and traditionally only show a single number. Gauge charts don’t align with lessons learned from Cleveland’s studies of graphical interpretation and precision.  People are really bad at measuring and comparing angles, so why not use something easier for people to understand like a bar graph? Gauge diagrams are simplistic, waste valuable space, and lack context. Look at the difference below between visualizing satisfaction interview data with gauge diagrams compared to bullet charts.  It is apparent that if you were putting this data in a dashboard, using the bullet chart both conserves space and allows for better data comparison.

With all that said, SOMETIMES they can be a good choice.  Let’s talk about two scenarios when I think they can be useful.

Scenario One: Visualizing the sweet spot

When you are visualizing a measure where the sweet spot is in the middle, a gauge diagram can be a great choice. For example, I am visualizing employees’ rating of their boss’s level of involvement in projects. Having a low score would mean their boss wasn’t involved enough; whereas having a high score would mean the boss was too involved and micro-managing. The sweet spot is in the middle. A gauge diagram is one of the only charts that can visualize this quality effectively. People hold a lot of assumptions when reading data from other graphs about slope and rate of change. Typically a high value is good or low value is good. Rarely are we asked to visualize the sweet spot and when we are, the gauge diagram is a solid choice.

Scenario Two: When a big, simple visual breaks up qualitative narrative

There are so many nuances to qualitative data that provide an opportunity for our audience to really get a deep understanding. This means that qualitative reports are often bogged down by long narrative. This is the perfect situation where we need a big visual to balance out all of those pages and pages of words. Sometimes we need a big, simple visual to give people’s eyes and brains a break.

Let’s say you are reporting client satisfaction across programs at your non-profit. You did some interviews with clients and categorized their overall sentiment and responses into three categories: not satisfied, satisfied, and very satisfied. You are giving the board a short qualitative report, but you want some simple way to get across what program is and isn’t working so they can allocate resources accordingly. The very thing most people don’t like about the gauge diagram – it’s waste of space – becomes an advantage when working with narrative-heavy context. Yes, gauge diagrams may not be efficient on a data dashboard – but this isn’t a data dashboard.  Take a look at the difference between these two reports.

Report 1 uses the space-efficient bullet graph:

Report 2 capitalizes on the large gauge visual to break up all the narrative:

Some guidance:

  1. Make gauge diagrams simple with no tick marks, extra decorative elements, or unnecessary clutter. Never put more than one dial in a diagram.

  1. Add a strong title for more bang for your buck.

  1. Add some context and raw data to back up your qualitative categorization.

  1. Use color intentionally to emphasize your point.

These two scenarios (with some suggested guidance) offer opportunities where a gauge diagram effectively visualizes qualitative data. Want to learn how to make one? We used this tutorial, which is based in Excel.

We are committed to contributing to the ongoing discussion and development of qualitative data visualization. The Qualitative Chart Chooser is still a living document. Keep us updated on what is working and what isn’t. I promise to continue to dive into detail on more qualitative visualizations in the months ahead.

Learn in the Academy!

You can find step-by-step instructions on how to make 60+ awesome visuals in my Evergreen Data Visualization Academy.

Video tutorials, worksheets, templates, fun, and a big-hearted super-supportive community. Learn Excel, Tableau, R or all three. Come join us.

Enrollment opens to a limited number of students only twice a year. Our next enrollment window opens April 1. Get on the wait list for access a week earlier than everyone else!

Master Dataviz with Graph Guides!

Our newest program, Graph Guides, is a custom-built, year-long sprint through 50 Academy tutorials.

When you enroll, we’ll assess your current data viz skill set, build you a customized learning path, and hold your hand as you blaze your way to new talents.

We open enrollment to 12 students at a time and only twice a year. Get on the waitlist for early access to our next enrollment window.

One Free Pie Chart

If I could, I’d ban pie charts. Not because they are inherently bad, but because most people don’t know the pie chart’s tiny range of acceptability. Here it is: Pie charts are ok if you don’t need the numbers on the wedges to get the story. That usually means you’ll be limited to very few slices. And you’ll need a great title. The largest slice starts at noon and the rest go in descending order, clockwise around the pie.

When I say these rules in my data visualization workshops, I see some people transcribe them in their notes, word for word. And I know these people are never, ever going to give up their pies. You probably have that colleague that just can’t quit the pie. Maybe its your boss and you hate to push back too much. Maybe its YOU.

So, just for you pie lovers, may I gift you with this pass for one pie chart:


I can respect a deep commitment, so have at it. But just once. One pie. One wacky pie chart among other really fantastic data visualizations will not hurt (too much).

Print this out and give it to your pie lover. Just right-click on the free pie card image, save it somewhere, and print.

How We Unintentionally Perpetuate Inequality Part 3

Last week a friend told me the story of a well-intentioned nonprofit that designed teaching materials for Native American populations, in which they unintentionally perpetuated inequality. He described one piece that included a picture of the scales of justice. I didn’t get the significance, so he explained that while most white people see the scales of justice as a symbol of fairness, the scales were used historically to rob Native Americans of their assets because white pioneers would use empty or extra heavy weights to tip the scales in their favor.

This white girl never knew about that symbolism. And I’d love to learn more. My Google searching isn’t turning up much (no surprise) so if you know of more reading I can do on this subject, please post about it in the comments.

The larger point here is that visuals matter to people. Visuals can carry deep cultural associations that many well-intentioned white people can miss. And we use them all the time in our reports and slides, even paired with a single large number as a way to visualize data. Crap! Wake up, Stephanie! This is a space where it hadn’t occurred to me the extent to which I need to listen to people of color and teach other white people how to be more thoughtful. So here are two stories that can educate us.

My colleague Beth was working on a report template for a client that would ultimately hold data-driven findings for Tribal grantees. She went to Shutterstock for some images. She said she searched that site using these search terms: Native American, American Indian, Tribal Nation, and Indian. Beth said, “I would not use the term ‘Indian’ to identify myself but it was a word used to describe the first people of this land.”

Her search results were not acceptable. These are what came up when she searched on “Native American.”

And these are what came back when she searched for “American Indian.” Beth put a blue slash through the ones she found offensive.

Beth said, “I wanted images of American Indian or Alaska Native people being physically active and I eventually found a few on Shutterstock after scrolling through many offensive images.”

My colleague Vidhya explains another example from an infographic on poverty:

“One bullet was about the extent to which people of color are more likely to experience poverty and beside that factoid, they inserted a photo of a smiling African American family as Exhibit A.” 

I don’t have the exact image Vidhya saw, but I’m sure you can imagine it, since that sort of imagery is everywhere.

Vidhya continues: “It was obvious that the audience in their mind’s eye is white—and that African American families (rather than inequitable systems) are the emblem of poverty. People of color are not necessarily African American, and the smiling faces smacks of minstrel imagery in which African Americans are either ignorant of or happy to be in the position they’re in. I imagine they were deliberate about choosing a happy, heteronormative family (man, woman, and child) rather than a sad or angry one, or an image of a single African American or a woman and child alone. I recognize that those would have probably been problematic in their own right, but that is part of the problem with a mindset in which African Americans are the image associated with a point that was really about racial disparities. For there to be a disparity, there has to be more than people of color involved. Whites get to be associated with regular old people, though they are just as implicated in the problem as people of color are IF we see the problem as inequity. If we insist on seeing the people of color experiencing poverty as the problem, no matter how sympathetic we may be to their plight, we face the conundrum of ‘the single story’ as Chimamanda Ngozi Adichie calls it.”

White people, we’d better put some more thought into this. Can you do your homework? Can you take the risk of running your imagery by a trusted friend or colleague and getting some feedback? Can you put yourself out there and test your work with some sample audience members? Can you ask your favorite stockphoto sites for more accurate and less offensive imagery? In a new/revived era where putting people of color and women into their “proper place” has become acceptable, we who are well-intentioned are going to have to work harder to make sure we are not unintentionally perpetuating inequality.

Vidhya helped me with parts 1 and 2 of this blog series, so read those too.

My 2016 Personal Annual Report

Every year I say this has been the best year ever and every year it is true. Daily, I’m grateful for work I love with amazing people. Thanks for being a part of this with me!

Since 2011, I’ve been creating a personal annual report. It’s a dashboard of sorts that tracks my key performance indicators in a visual way.

I wrote yet another book this year! Our Chart Chooser Cards Kickstarter campaign reached 1,000% of our goal! Wow! And thank you.

Two Alternatives to Using a Second Y-Axis

Almost as often as I see a pie chart with a hundred tiny slivers, I see line graphs using two y-axes. And it is just as bad.


Graphs like this appear in every industry, everywhere I consult all around the globe. Using two y-axes is not a great idea because it gets confusing, fast. Which line goes with which dataset? At least here I have color coded the line to its axis, a smart improvement on what I typically see. Why do people insist on using a second y-axis? I suspect the answer is because they are just easy to make but I often hear reasoning like:

“I want people to see the relationship between these two things.”

Your heart is in the right place, darling, but putting them in the same graph introduces confusion points, like where the academic and behavior lines cross in the graph above. People see that intersection point and think it means something but it really doesn’t. Like “oh I wonder what happened that made behavior suddenly drop below academics.” But these two things aren’t on the same scale, so the point at which they cross is meaningless. By putting these two variables into the same graph, it implies more relationship between the two things being graphed than actually exists in real life.

“I need to fit it all into one graph.”

As in, there’s only room on the slide for one graph, huh buddy? That’s an unfortunate parameter to work within but both of the alternatives here would take up the same amount of space.


Alternative 1: Two Side by Side Graphs

Perhaps even easier than making one graph with two y-axes would be to make two separate graphs. Just nestle them up next to each other like best friends. We still see that academics are increasing while behavior is decreasing, without the complicating intersection we had when they were on top of each other. I inserted a textbox that spans across both graphs for the title.  Both graphs are not as wide as a default graph, so I can fit them into the same amount of space as the original graph with two y-axes. This alternative also lets me move the axes titles to a subtitle location, getting rid of the need to read lengthy text vertically.

Alternative 2: Connected Scatterplot

If you think about it, what’s really happening with two y-axes in the same graph is that we are trying to show a relationship between two continuous variables. That should ring a bell from grad school days. Two continuous variables are usually graphed as a scatterplot. Yet this data also runs over time – which makes a connected scatterplot a sweet, elegant answer. The line connecting the dots in the scatterplot suggests change over time. I added in the years as markers along the line so the chronology is clear.

At first, this graph is not the most intuitive to read. But it’s storytelling powers are pretty impressive. The first segment of the line shows that between 2011 and 2012, academic performance increased (the line moves to the right) but so did the number of behavioral referrals (the line goes up). The second segment continues the story: between 2012 and 2013, academic progress continued (the line goes right) but schools got behavioral referrals under control (the line goes down). This is one detailed story we can see by tracing a simple line. And we can always annotate that story right within the graph using a textbox, like I did here.



You’ve got two alternatives here, one simple option and one that’s more sophisticated. Either one is going to be a win over the traditional double y-axis graph and both alternatives can still be made right inside Excel. Time to print out this blog post and slip it anonymously into your colleagues’ mailbox!

Data Nerd Holiday Swag

One of the smartest branding moves is to keep yourself top-of-mind with clients when you aren’t in the throes of discussing contract details. The new year is a perfect opportunity to remind your clients that you are both fun AND serious about data. I send a holiday card to my clients every year and I release my designs for you because I love you and I care about your work.
This is the holiday card I sent last year. Click on the picture to head to the order page.
And here are my cards from holidays gone by. Click a picture to head to the order page.

These tried and true designs have been bought by the bucketload by people just like you and sent to happy clients who now see the sender as charming and thoughtful. Just the way to start a new year.

Slopegraph Holiday Card for Data Nerds
All I Want Holiday Card
Happy Holidays Data Nerd Card
Prosperity Pie Chart Holiday Card
Maybe your branding needs to happen on a more personal level. Maybe you need to remind your data nerd sweetheart that you support their full nerd power, even though you quibble about the shopping list and stress out over the holiday calendar. Warm your data nerd’s heart with a flask. BAM shopping done. Both designs says “DRINKS FOR DATA NERDS.” They have a y-axis which is the amount of liquid remaining in the flask and an x-axis called “my condition.” The difference is the data points. They both cracked me up so I just made two designs for you to choose from. Bonus points for gifting it full of booze. Click a picture to head to the order page.
Drinks for Data Nerds Flask 2
Drinks for Data Nerds Flask 1
Despite how 2016 has been one heck of a year in so many regards, we can still spread love and goodwill. Thanks for being a bright spot for me this year.

PS. Members of my Data Visualization Academy are getting discounts on this swag, in addition to video tutorials on advanced graph making, templates, coaching, and R code. Enrollment for another cohort of 100 is now open. Maybe *that’s* the holiday gift your favorite data nerd really needs.

Qualitative Chart Chooser


Hey there.

Having a rough day?

I have something to make it a little brighter.

A Qualitative Chart Chooser. 

Actually, you can find our most recent free version here.

And the most up to date version in Chapter 8 of Effective Data Visualization, where we have the largest compendium of qualitative viz options available.

We teach you how to make these in Excel, Tableau and R as part of the Evergreen Data Visualization Academy. Come join us.


Visualizing qualitative data is like making homemade risotto. You are standing over the stove (aka hunkered down with your computer), waiting patiently for the magic to happen. It’s slow and sweaty, but in the end SO worth it. There is a reason you can’t order risotto at McDonalds, and there is a reason you can’t display your qualitative findings in a nice neat dot plot. I am going to share some resources and ideas that will help give your audience a taste of your rich qualitative findings.

The reality is, most people are never going to be excited to read your text heavy 50-page report with no visuals. This is where data visualization can come in handy. Visualization is a great tool to get people interested and engaged with your story. The problem is, many of the qualitative visualizations I see are reports with endless callout quotes or ugly charts that were spit out of data analysis software. We can do better than this.  Let’s explore our options.

Take a peek at the qualitative chart chooser(s) we made. This is an attempt to organize different ways to show qualitative data. The truth is, this was hard. We have gone through many revisions. Like qualitative data, categorizing the options wasn’t a straightforward process. I am going to share with you the two best drafts we came up with. Here at Evergreen Data, we like to start with your story. Your message should be the foundation of all explanatory qualitative visualizations.

Draft One:

qualitative-chooser-1-0This is the draft we handed out during a qualitative panel at the American Evaluation Association conference. The different kinds of stories you tell with your data are in grey boxes on left. To the right are different chart options that help tell that story. This is all fine and dandy, but things are not as straightforward as they look. There are lots of overlap. For example:

Icons can be symbolic and help categorize themes.


They can also help show alignment with a goal or outcome.


Because of this overlap and my difficulty with fitting the visuals into boxes, we went in a different direction with the second draft.

Draft Two:

This one is two pages. The first page has the options categorized by visualization type. Because this doesn’t go into detail on the story each visual tells, we included a matrix that does. The matrix shows the interconnectedness and complexity of visualizing qualitative data.

qualitative-chooser-2-0_page_1 qualitative-chooser-2-0_page_2

Let’s look at an example of when the chart chooser can be helpful. I am evaluating the factors that impact work culture at a local community healthcare center. Based off the key informant interviews, I found two themes: cross collaboration and connectedness. The client is particularly interested in differences between staff. During data analysis, I found that within these themes, management staff saw collaboration as the most important factor influencing work culture. Program level staff, on the other hand, found connectedness to be most important. What is the best way to display this difference? A spectrum display is a good fit because it helps show themes by quantifying individual cases centered around a mutually exclusive variable.

Stuart Henderson does a great job describing and analyzing this visualization in his article “Visualizing Qualitative Data in Evaluation Research” in AEA’s journal New Directions for Evaluation. A spectrum display compares the relationship between qualitative cases and themes. You must have a mutually exclusive variable. In the example below, 12 key informant interviews were conducted with staff. The mutually exclusive variable in this display is staff type. The two themes are displayed at the bottom of the spectrum. Each case is coded on the presence of that theme.


At first, this display comes off complex and overwhelming. It is hard to tell that my main point is to show differences between staff perception. Choosing the best visual for your message is just the first step. We can apply data visualization techniques to transform our visual into something useful. After breaking the visual into small multiples, using strong titles and color coding, and adding quotes, the message becomes much clearer.


My biggest piece of advice is to stay true to the data. Whenever possible, link a visual with quotes and narrative that help provide evidence and context to your main point.

Learn in the Academy!

You can find step-by-step instructions on how to make 60+ awesome visuals in my Evergreen Data Visualization Academy.

Video tutorials, worksheets, templates, fun, and a big-hearted super-supportive community. Learn Excel, Tableau, R or all three. Come join us.

Enrollment opens to a limited number of students only twice a year. Our next enrollment window opens April 1. Get on the wait list for access a week earlier than everyone else!

Master Dataviz with Graph Guides!

Our newest program, Graph Guides, is a custom-built, year-long sprint through 50 Academy tutorials.

When you enroll, we’ll assess your current data viz skill set, build you a customized learning path, and hold your hand as you blaze your way to new talents.

We open enrollment to 12 students at a time and only twice a year. Get on the waitlist for early access to our next enrollment window.

Election Night Trap Map

Since none of us can escape election coverage, no doubt you’ve seen voter and poll data displayed on some unusual maps. 538 is showing electoral votes as a hex map (made up of hexagons) and the Wall Street Journal did the same as a tile map (made up of tiles, or squares). The general idea is that our traditional map gives a lot of visual weight to places that are geographically large, like Montana, even though there are few people and electoral votes there. By giving states the same size, visual interpretation is a little easier.

I created a Trap Map, where each state is a trapezoid. Also because Trap Map is a pretty cool name.


At my house, we have printed several trap maps and gathered our red and blue crayons to keep track as the election results roll in tomorrow night.

You can play along too. Download the Trap Map. I hear coloring eases anxiety.


Updated Data Visualization Checklist

WHOA THERE: This checklist is now a website where you can upload your image and we will walk you through each checkpoint, helping you rate yourself.

Tweaked and clarified, here is the updated Data Visualization Checklist.


Ann and I adjusted 5 of the checkpoints.

Text size is hierarchical and readable

Titles are in a larger size than subtitles or annotations, which are larger than labels, which are larger than axis labels, which are larger than source information. The smallest text – axis labels – are at least 9 point font size on paper, at least 20 on screen. We updated this description to talk about what size goes where.
Here’s an example:
Both text size and use of shades of gray indicate the hierarchy of information, with the title being in the largest and darkest and the source information the lightest and smallest.

Labels are used sparingly

Focus attention by removing the redundancy. For example, in line charts, label every other year on an axis. Do not add numeric labels *and* use a y-axis scale, since this is redundant. We updated this description to discuss how one should choose either gridlines or labels. A terrifically bad example:
So much redundancy happening in this graph, but Mrs. Glosser definitely doesn’t need a y-axis with all those gridlines and the exact number labels on each marker. If the exact values are important, directly label. If the overall pattern is sufficient, use the y-axis.

Proportions are accurate

A viewer should be able measure the length or area of the graph with a ruler and find that it matches the relationship in the underlying data. Y-axis scales should be appropriate. Bar charts start axes at 0. Other graphs can have a minimum and maximum scale that reflects what should be an accurate interpretation of the data (e.g., the stock market ticker should not start at 0 or we won’t see a meaningful pattern). We updated this description to address the y axis debate.
This example should have an axis that starts at zero. (Chris Lysy calls this the Cable News Axis.)
But in this example, a y-axis that starts at zero would wash away all ability to interpret what’s important:
In cases like the stock market, zero would never be within the set of possible values (god forbid) so it doesn’t make sense to include it in the axis. For your own data, you might consider a minimum based on historic lows and a maximum based on your goal.

Axes do not have unnecessary tick marks or axis lines

Tick marks can be useful in line graphs (to demarcate each point in time along the y-axis) but are unnecessary in most other graph types. Remove axes lines whenever possible. We updated this description to say axes lines should be removed when possible.
Check out how this example uses no axis lines at all, which produces a cleaner graph:
A vertical axis line would have put an unnecessary division between the category label and it’s data, which doesn’t make sense. A horizontal axis line would have been useless.

Graph has appropriate level of precision

Use a level of precision that meets your audiences’ needs. Few numeric labels need decimal places, unless you are speaking with academic peers. Charts intended for public consumption rarely need p values listed. We updated this description to talk about when to use decimals or show p values.

This example from Brookings was posted to their blog, meaning the audience was a public one:

precisionexampleBut note the asterisks with the 3 different levels of p. Few people in a public audience even know what p means and those people don’t care about the 3 different levels. Now, Brookings’ economics peers will care a lot about that sort of thing (but not the stupid sun in the background). Level of precision is audience dependent. (And don’t get me started on standardized coefficient value).

We have some super fun developments of the Checklist coming in the future but for now, download the updated Data Visualization Checklist and let it guide your graph development.

Learn in the Academy!

You can find step-by-step instructions on how to make 60+ awesome visuals in my Evergreen Data Visualization Academy.

Video tutorials, worksheets, templates, fun, and a big-hearted super-supportive community. Learn Excel, Tableau, R or all three. Come join us.

Enrollment opens to a limited number of students only twice a year. Our next enrollment window opens April 1. Get on the wait list for access a week earlier than everyone else!

Master Dataviz with Graph Guides!

Our newest program, Graph Guides, is a custom-built, year-long sprint through 50 Academy tutorials.

When you enroll, we’ll assess your current data viz skill set, build you a customized learning path, and hold your hand as you blaze your way to new talents.

We open enrollment to 12 students at a time and only twice a year. Get on the waitlist for early access to our next enrollment window.

Announcing Chart Chooser Cards

Update: After a successful Kickstarter campaign where we raised over 1,000% of our goal, the cards are in production and you can now order a deck, an infographic, and our templates from our permanent website. Thanks for your support, lovely people.


Chart Chooser cards are simple and easy to use. They help you choose the best type of chart to display and format your data.


Each chart card shows you the common name of the chart type, a description, a visual example, when it is used, and what type of data set it’s best for.

The chart chooser deck has 51 cards, including 33 chart types organized into 6 data categories to help you quickly sort through the noise. 13 bonus cards help you sort through what people are looking for in data, and how to format your chart.


I’ve been using these cards in my data visualization workshops. They are an awesome tool to narrow down the wide world of chart choices and identify the right one for your data.

Want a deck? The cards are now available as part of a Kickstarter campaign. We offer different pricing levels with all kinds of awesome swag, including:

An infographic of the chart types available in the Chart Chooser Cards

An Excel file with every single graph within the deck, already made and ready for your data

A Tableau file with every single graph within the deck (Bless you, Andy Kriebel!)

One-year of access to my currently-closed-to-enrollment Data Visualization Academy

With all of these tools as your fingertips, you’re well on your way to being a Data Visualization Rockstar! Check out your options on the Kickstarter page.

From the blog