### Overview

Several weeks ago the data visualization community broke into justified outrage over an inexcusably misleading dual-axis chart from Americans United for Life.  I plan to write an article about this and other “ethically wrong” visualizations in a few weeks but in the meantime I encourage you to read these excellent posts from Alberto Cairo and Emily Schuch, as well as this discussion from Politifact.

Around the same time these posts appeared I came across a “Viz of the Day” dashboard from Emily Le Coz that accompanied a lengthy article in the Daytona Beach News-Journal.  The dashboard contained several visualizations but the one that caught my eye was this dual axis chart.

Figure 1 — Infographic showing that as the number of firefighters has increased over the past 30 years, the number of fire-related deaths has decreased.

I engaged in an interesting Twitter discussion about this graphic with Alberto Cairo, Jorge Camoes, and Noah Illinsky. I’ll get into that discussion in a bit (and point out some troubling problems with the visualization) but first want to discuss the use case for dual axis charts.

### Why use dual axis charts

There are several reasons to use a dual axis chart (e.g., a Pareto chart that shows individual values along with the cumulative percent) but the primary use case is when you want to compare two completely different measures and see if there is any noteworthy relationship between the two measures.  Consider the example below that shows cyclical sales data for a retail store (bars) and the number of orders placed each month (line).

Figure 2 — Dual axis chart comparing sales and orders by month.

The surprising result is that while November is historically the strongest month for sales (\$5M from 2010 to 2013) the total number of orders placed in November is the lowest of any month. And yes, I checked to make sure that this was true of all years and not one crazy blowout year.

I think this dual axis combination chart (where we show bars and a line) makes it easy to see there is something very interesting about November. The low number of orders combined with the high sales – something that is easy to see – means that we either sold more items per order or more expensive items per order.

### So, what’s wrong with the firefighter example?

Given that dual axis charts can be so useful I wondered why I had problems with the Firefighter example.  Fortunately, the author made the dashboard downloadable from Tableau public so I was able to see how it was put together.

#### Cutesy icons set the wrong tone for the piece

My first problem was with the firefighter hat and skull-and-crossbones icons.

Figure 3 — Icons representing firefighters and civilian deaths.

In my opinion (and it is just an opinion) I thought this “cartoonified” the visualization. I would much prefer to see either a simple color legend or a label next to both lines.

#### The author exaggerates the changes over time

A much more troubling issue is that the author uses a fixed Y-axis that exaggerates the changes over time.  The author also fails to show the axis labels so we can’t see that the axis doesn’t start at zero.

Consider the dashboard below that shows the original visualization on the left with an accurate visualization on the right.

Figure 4 — Comparison of fixed axis vs. automatic axis charts.  Note that the axis uses a SUM() function while the label is using AVERAGE(). The data is repeated three times in the data source which is why the author needs to use AVERAGE(). Yes, the axis should use AVERAGE() as well but the relative positioning of the elements is the same with SUM() so this causes no harm.

Because the author fixed the Y-axis rather than starting from zero, the slope of the lines is exaggerated. While this does not alter what is in fact a noteworthy observation, whenever I see this type of “rigging” it makes me question the validity of any and all parts of the story.  That is, even though I don’t think the exaggeration was an intentional attempt to dramatize the difference, seeing this in play will make me question everything that the author and the publication now publishes.

Am I being too hard on the author? I don’t think so as anything that’s published as a “viz of the day” and accompanies a high-profile news article should get a lot more scrutiny than just any old Tableau Public visualization.  While I don’t feel mislead by the overstated changes, I do wonder at what point does a viz cross the line into TURD territory (Truly Unfortunate Representation of Data)? We’ll save that discussion for a later post.

### Different approaches

#### Combination area and line chart

After adjusting the axis I still wondered if having two line charts was causing unnecessary confusion. In my first makeover attempt I tried combining an area graph with a line chart, as shown here.

Figure 5 — First makeover attempt.  A dual axis chart using an area chart for firefighters and a line chart for civilian deaths.

While using two different chart types made it easier to see that I was comparing two different measures, I didn’t love the chart and sought alternatives.

#### Connected Scatterplots

On Twitter Jorge Camoes offered this connected scatterplot.

Figure 6 — Jorge Camoes’ connected scatterplot.  Notice that the axes do not start at zero but that the axes labels are at least visible.

In a connected scatterplot the path the line takes represents the year.  This is why the line folds back on itself from time to time (more on this in a moment).  Camoes also “normalized” the data using an index so that both civilian deaths and number of firefighters start at a value of 100.

I like this visualization very much but fear that many people won’t understand the index value of 100 so I tried my own connected scatterplot, shown below.

Figure 7 — Connected scatterplot with regular vs. normalized values.  Notice that the X-axis does not start at zero but that the axes labels are visible.

Before anyone cries foul about the X-axis, here’s a version with the axis starting at zero.

Figure 8 — Connected scatterplot with both axes starting at zero.  This may be why Camoes normalized the data although his chart doesn’t start at zero, either.

I think starting the x-axis at zero obscures the relationship but that’s not what makes me question using this approach.  My problem is that many people will have a hard time understanding how the line “works”, as it were.  This is because whenever we see a line chart that involves time we come to expect marks on the left of the chart to show older dates and marks on the right to show newer dates.  In other words, we expect the chart to behave like this.

Figure 9 – Since grade school we’ve been indoctrinated to expect earlier dates to the left and later dates to the right.

With a connected scatterplot the X-axis is “owned” by an independent measure so we have to adjust our perception to see that sometimes a later year will appear to the left of an earlier year, as shown below.

Figure 10 — Connected scatterplot with marks showing all years.

Notice how 1986 appears to the left of 1985 and 1989 appears to the left of 1988.  Unless you are used to this type of approach this can look very strange.

### Keep it simple

After experimenting a bit more I decided to forgo the dual axis and connected scatterplots and fashioned this simpler narrative.

Figure 11 — Two separate charts yielding a simple and easy-to-follow narrative.

If you have what you think is a better approach I would love to see it.  If you’re using Tableau you can download the packaged workbook with the original dashboard and various makeover attempts here.

### Overview

I recently wrote about emotional vs. accurate comparisons and several people questioned whether the word “emotional” was appropriate.  (Several people questioned my assertions, too.  You can read their comments here.)

For this discussion I’ll use the term “engagement” in place of “emotion” and we’ll look into the challenges of creating public-facing visualizations that attract and engage, are clear and accurate, and do these things without “dumbing down” the subject matter.

### Time Magazine and a cumbersome infographic

Stephen Few recently wrote a great post about the following infographic that appeared in Time Magazine in August, 2015.

Figure 1 — Time Magazine’s “Why we still need women’s equality day” infographic. See http://time.com/4010645/womens-equality-day/.

I have three major problems with this treatment.

1. This is an important subject but the cutesy approach trivializes it.
2. With so many chart types I have to work very hard to make comparisons among the different areas (Federal, Congressional, etc.). In addition, the chart is very long and requires a lot of scrolling.
3. I strongly suspect that most people thought this was a dashboard having to do with Republicans and Democrats. I know that for me, whenever I see red and blue in a political context I think Republicans and Democrats and I had to fight this expectation to see that this was about men and women.

### Stephen Few’s redesign

Here is Few’s redesign.

Figure 2 – Stephen Few’s clear and compact redesign.

The collection of stacked bars makes it very simple to compare across the various categories and treats an important subject with the seriousness that is warranted.

But…

Few’s treatment is rather clinical and may be a little too dry for Time Magazine.

So, is there a way to fashion a graphic that is clear and accurate, like Few’s, but does more to draw the reader in?

### Alberto Cairo’s redesign

Stephen Few asked Alberto Cairo to have a look at the source graphic and Cairo was able to turn out the following in a matter of minutes.

Figure 3 — Cairo’s redesign of Few’s redesign.

Here are Stephen Few’s comments upon seeing the redesign:

“Alberto,

You’re the man! I love your improvements to the graphic.

You described your version as middle ground between my position and that of the embellishers, but I don’t see it that way. I’m an advocate of the kinds of embellishments that you added to the graphic for journalistic purposes, for they don’t detract from the information in any way. I’ve always said that journalistic infographics can be both informative and beautiful without compromising either. Doing this takes skill, however, that relatively few of the folks producing infographics possess. It also takes graphic design skill that I don’t possess, which is why I don’t design journalistic infographics. You’ve illustrated what it takes to do this well. As I said, you’re the man.”

I think Cairo would be the first to agree that there are many shortcomings to his rendering (e.g., colors, the guy on right looks like he’s holding a boomerang and not reading a book, etc.) but remember, Cairo put this together in a few minutes simply to show that it is in fact possible to create something that is beautiful and emotionally engaging without sacrificing one pixel of analytic integrity.

Tagged with: , , , , ,

### Overview

I’ve conducted a lot of Tableau training classes and have found three things that confuse students simply because of the nomenclature Tableau uses for these things.  These three terms are

• Table Calculations
• Quick Filters

Consider the chart below that has both mark labels and an axis along the bottom.

Figure 1 — Bar chart with visible axis.

Because each bar has a label we don’t need to see the axis.  We can hide the axis by right-clicking it and selecting…

Figure 2 — Turning off the header turns off the… footer.

Yes, indicating that we don’t want to display a header will make Tableau hide…

the footer!

As I explain to students, in Tableau anything that surrounds a chart is called a Header.  If it’s along the top of a chart, it’s a Header.  Left side of the chart?  Header?  Bottom?  Header.  Right side?

### Table Calculations

I know the first time I saw this I thought “Table Calculations” pertained to a visualization that used text tables. As I explain to students, I think of table calculations as Tableau having the ability to do math in its head.

Consider the example below where we show the raw vote count for each candidate from the 2012 US presidential election.

Figure 3 — Bar chart based on query to the source database

Here, Tableau has queried the underlying database and is displaying the results based on that query.

With a table calculation, Tableau looks at the results that are already on display, as it were, and then does some additional internal calculations.  In the case of asking Tableau to show the percent of total, Tableau adds up the total for all three candidates and then divides the tally for each candidate by that total.

As I said, I find it helpful to think of Tableau Calculations as Tableau doing math in its head.

### Quick Filters

To filter results in Tableau, you drag dimensions and measures from the Data window to the Filters card and then apply the settings you want for the various filters.

If you want easier access to the filter settings you can right-click a filter and select Show Quick Filter.

The problem with this term is that people new to Tableau think this pertains to speeding up the filter when it in fact means that you just want the filter control to be visible on a worksheet or a dashboard.  It has nothing to do with making filters quick.  In fact, having lots of quick filters on a worksheet can slow Tableau down because Tableau has to calculate what selections should appear in each of the quick filters.

The only rationale I can see for the name is that it allows you to access the settings quickly rather than having to go through the Filters dialog box.  Still, it’s quite confusing for those first learning Tableau.

### Summary of confusing terms

Here’s a summary of the terms that often confuse people new to Tableau.

 Term What students think it means What it actually means Header Something at the top of a chart Anything that surrounds a chart Table Calculation Something having to do with text tables / cross tabs The ability for Tableau to do math “in its head” Quick Filters Some setting that makes filters work faster Make the filter control visible

### What should we call these things and should Tableau rename them?

Given just how entrenched Tableau is it may be too late to change these terms, but if it’s not too late…

In the case of Show Quick Filters I would change it to Show Filter Control.

Tagged with: , , , , , ,

### Overview

Figure 1 – Bar charts are better than pie charts are better than donut charts.  Most of the time.

As anyone who has read this blog knows I’m definitely a “bar charts are better than pie charts are better than donut charts” kind of guy, at least when you need to make an accurate comparison.

But in my classes, as I rearticulate the case against pies and donuts, I find myself wondering if there are in fact times when a pie chart might be a better choice.

Most of my data visualization work is for internal purposes so I focus on making it easy for people to make an accurate comparison.

But as my clients and I make occasional forays into public-facing visualizations I think about how to make it easy for people to make an emotional comparison.  By this I mean that I want people viewing the visualization to just “get it”.

Better yet, I want people to get it, be engaged by it, and in some cases, “feel” it.

With this in mind, in this post we’ll explore cases where

• a pie chart is in fact as good, if not better, than a bar chart.
• circles and spheres do a better job conveying magnitude than do bars.
• a waffle chart produces an emotional wallop without compromising analytic integrity.

### Where a pie chart trumps a bar chart

So, it’s the year 2034 and in this somewhat dystopian future there’s a movement afoot to add an amendment to the US constitution banning the use of pie charts.

Those of you familiar with the United States Constitution know that three-quarters of the states need to approve an amendment for said amendment to become law.  In 2034 it turns out the 39 of 50 states will in fact ratify the amendment.

Does that get us the needed 75%?  Here’s a simple, compact chart that lets us know immediately.

Figure 2 — The amendment banning pie charts passes as I can see that the “Yes” votes fill more than three quarters of the circle.

It’s so easy to see that the “Yes” votes fill more than three-quarters of the pie that I don’t need labels indicating the large slice is 78% and small slice is 22%.

Compare this with a bar chart.

Figure 3 — Did the “Yes” exceed 75%?  Without labels it’s very hard to tell.

Without labels showing the percentages I cannot tell for sure if the “Yes” bar is more than three times larger than the “No” bar.

Okay, Okay, Okay!  I know that a simplified bullet chart would work, too.

Figure 4 — A bullet chart shows that we’ve exceeded the goal.

Yes, the bullet chart makes it clear that I’ve exceeded my goal but I need to know that the goal was 75%.  I don’t need the goal line with the pie chart.

So, does this mean that it’s okay to use pie charts instead of bar charts?

No.  Based on this example it’s only okay to use a pie chart (singular).  In addition, your pie chart (singular) needs to meet the following conditions:

• One of the slices has to make up at least 50% of the pie.
• If you’re pie has more than two slices you don’t ask people to compare the smaller slices.

### Where circles and sphere’s do better than bars

As we all know Jupiter is big, really big.

Just how much bigger is it than Earth?

Should I create a bar chart to show this? If I were to create one should I compare the radius or the surface area of each planet?

Or should I really go nuts and compare the volume of the planets?

I don’t think the dashboard shown above is nearly as effective as the visualization shown below.

Figure 5  — “Size planets comparison” by Lsmpascal – Own work. Licensed under CC BY-SA 3.0 via Commons – https://commons.wikimedia.org/wiki/File:Size_planets_comparison.jpg#/media/File:Size_planets_comparison.jpg

Jupiter and Saturn – and even Neptune and Uranus – really dwarf earth and the other planets and with this visualization I feel it.

Even the simple chart comparing the area of the cross section of the planets gives me a better feel for the data than does the bar chart.

Figure 6 — Circles comparing cross-section area of the planets.  Yup, I can tell that Jupiter is way bigger than Earth.

Is it essential that I can tell exactly how much larger one planet is than another?  I don’t think it is and I much prefer the emotional pull of the circles and the spheres.

#### A Fun Tangent

One thing that’s very hard to express in a static chart is how much space there is between the sun and the planets.  To get a sense of just how incredibly vast the distances are check out this fascinating, albeit somewhat tedious, interactive visualization from Josh Worth.

### Getting an emotional wallop with waffles

A few weeks ago Cole Nussbaumer posted a tweet asking people what they thought of this chart from The Economist:

Figure 7 – A waffle chart from the article “Teens in Syria”.  See http://www.economist.com/blogs/graphicdetail/2015/08/daily-chart-6?fsrc=rss.

The first thing that surprises me about this is that The Economist went with a waffle chart and not a bar chart, like the one below.

Figure 8 — The type of chart I would have expected to see in The Economist.

The second thing that surprised me was that I preferred the waffle chart.  Yes, as Jeffrey Shaffer correctly points out, the dots are so tightly packed that you literally see stars between the circles, but  this can easily be remedied.  The question on my mind is why do I prefer waffles?

My answer is that the having each dot represent one of the 120 people surveyed connected with me in a way that the bar chart did not. Combined with the percentage labels (which are critical to the success of the visualization) the waffle chart hit me hard and it did so without dumbing down the importance of the discussion one bit.

### So, are bars charts always boring?

No!  In my next blog post I’ll show you an example of a bar chart embedded inside a “come hither” graphic that

• attracts and engages
• does not trivialize an important issue
• represents the data clearly and accurately

Stay tuned.

# Overview

I’ll admit that I have a problem with treemaps in Tableau, but it’s not because the chart type is in some way inferior. My problem is with how people use – and misuse – treemaps.

Here’s a good example of misuse.  Instead of displaying something straightforward that looks like this…

Figure 1 — The humble, but accurate bar chart

… some people feel compelled to add “visual variety” to their dashboards and instead create something that looks like this.

Figure 2 — Look , Ma! I made a Mondrian!

Except for the “it looks cool” factor there’s no good reason to use a treemap in this situation.

So, when should you use a treemap?

# What’s in a treemap and why it can be useful

With a treemap you have two attributes at your disposal:

1. The size (area) of rectangles, and
2. The color of the rectangles

A treemap consists of packed rectangles where the area of a rectangle corresponds to the size of a particular measure.  In the example above the size of the rectangle is based on the number of people that come from a particular region.  North America has the largest value so it’s represented with the largest rectangle. Europe has a smaller value to its rectangle is proportionally smaller.

Treemaps really come in handy is when you have A LOT of marks to plot and you need to show all of the marks in a compact area.

So, this sounds like a great chart – we’ve got rectangles to show how big and small stuff is, color to group related rectangles intelligently, and we can fit a lot of stuff in small space.  Why not use this chart all the time?

The downside is that we are comparing the area of rectangles and with rectangles it is difficult to make an accurate comparison. People may be very good at comparing the length of bars but as a species we are not particularly good at comparing the area of rectangles (and we’re downright awful at comparing the area of circles.)

So, given the advantages and shortcomings, just when should you use it?  Let’s look at a particular scenario.

# Showing Presidential Electoral Results

### A Filled Map

Consider the electoral map below showing electoral votes by state for Barack Obama and Mitt Romney in 2012.

Figure 3 — Filled map showing electoral votes for the 2012 presidential election (displaying 48 out of 50 states)

Our Electoral College system is fairly confusing and I can only imagine how somebody from outside the US would look at this as there appears to be more red on the map than blue… but the blue guy won!

This discrepancy becomes even more pronounced when we include Alaska and Hawaii in the map.

Figure 4 — Filled map showing electoral vote winners for the 2012 presidential election (displaying 50 states)

Clearly, a map designed to show how much area there is in a state fails with Electoral College results where the numbers are based on population not land mass.  In the example above there’s A LOT more red then blue, but again, the blue guy won the election.

Perhaps a different type of chart will do a better job?

### Symbol Map

Here’s a symbol map of the same data.

Figure 5 — Symbol map showing electoral vote winners for the 2012 presidential election (displaying 48 out of 50 states)

I think this is more accurate as there’s clearly more blue than red, but it’s still a tough read.  What else might work?

### Cartogram

Here’s a cartogram from Professor Mark Newman of the University of Michigan showing the same data, except the polygons for each state has been adjusted to reflect the population of the state.

Figure 6 — Cartogram showing election results where the shape of the state is based on its population and not land mass.

While it’s very clear that there is more blue than red on this map there are two problems with this approach:

1. There aren’t many tools that will support this type of distortion; and,
2. This map will frighten small children.

### Summary Bar Chart

Why not just display a simple bar chart showing the total number of electoral votes, like the one shown here?

Figure 7 — Electoral vote count by candidate

This is certainly very clear and we can see easily by how much Obama won, but we’re missing an important part of the story.

In US presidential elections a winner is chosen by tallying the electoral votes from each state and the summary bar chart doesn’t show us how each state contributes to the total for each candidate.

# And the Winner is… ? The Treemap!

Here’s a treemap showing the exact same data.

Figure 8 — Treemap showing 2012 electoral vote results

Of all the single visualizations I think this treemap tells the most complete story.  We can see just how much states like California, Texas, Florida, and New York contribute to the total as well as gauge —  to some degree  — just how many more electoral votes Obama received than did Romney.

One shortcoming, however, is that we can’t see the names of all the states as some of the rectangles are too small.

One way to address this is by adding a tool tip, as shown here.

Figure 9 — Hovering over a mark allows me to see the name of the state and number of electoral votes.

While this works, a problem we should address is that the small states are not easily searchable.  That is, if I want to know the results for Alaska, Hawaii, Delaware, etc., I have to go hunting for them.

At this point we’ve gotten about as far as we can get with a single chart.  To tell the complete story – and to make it easy for people to find results for a particular state – we should create a dashboard.

## The Electoral Vote Dashboard

Here’s a dashboard that puts two of the views together and that allows the user to find a particular state’s rectangle by selecting the state from a list.

Figure 10 — Electoral votes dashboard.  Selecting a state from the list will display that state’s rectangle in the treemap.

While the “star” of the dashboard is the treemap, the summary bar chart and the selectable list make the story complete and we get a solid understanding of the 2012 Electoral College results.

And we achieved this without using an actual map.

Tagged with: , ,

## Overview

So, here’s why until recently I’ve recommended that my clients avoid large dashboards.

We’ve been working on a collection of killer dashboards and we’re all set to make a big presentation to the CEO. This thing is so high profile we get to use the executive conference room with the super bright projector and the 120-inch screen.

Our dashboards are all 1,325 x 1,000 pixels, but they’re going to look fantastic on that giant screen.

We’re incredibly well prepared.

At least we think we’re incredibly well prepared because when we arrive an hour early we discover the top resolution of that ever-so-fancy projector is 1,280 x 800 and our ever-so-well-crafted dashboards won’t fit on the screen.

It doesn’t fit! Tableau Desktop and Reader will not scale the dashboard intelligently and we end up with the dreaded scroll bars.

Yikes, we have scroll bars! What are we going to do?

And don’t suggest using Tableau’s “Automatic” dashboard setting as it will just squish the different visualizations and won’t scale the fonts.

## Let your browser scale the dashboard

While Tableau Desktop and Reader cannot scale your dashboard, Tableau Public, Tableau Online, and Tableau Server — with the help of your browser — can scale the dashboard, and scale it intelligently.

For example, using Tableau Public with the  “Zoom” feature in Google Chrome…

… allows us to “fit” the dashboard on our large, but relatively low-resolution, screen.

It fits! Thank you, browser.

## Conclusion

If you are presenting your work using Tableau Desktop or Tableau Reader then you either have to compose for the lowest-common-denominator screen or live with scroll bars.

If, however, Tableau Public, Tableau Online, or Tableau Server are an option, you should be able to use your browser’s zoom feature to make sure your dashboards fit on the screen.

Tagged with: , , , , ,

## Overview

Addressing this scenario has been the most popular request I’ve received over the past year. Here’s a summary what my clients and students have asked:

• How do I show the change in Sales, Percentage of Promoters, Number of Visits, etc., between this month / quarter / year, and the previous month / quarter / year?
• How do I make it easy to see which areas of the organization had an increase this period and which had a decrease?
• How do I make it easy to see how much greater / less this period’s numbers are than the previous period?
• How do I determine and show if this change is statistically significant? That is, how do I apply the stat test we like to use in our organization?
• If the change is statistically significant, is it a one-time thing or should I start hyperventilating?

This is a LOT to take on and we won’t be able to fit all of it into a single visualization.

But we can fit it into a compact dashboard.

## Important Ground Rules

In the example that follows I look at the percentage of people that responded with a “9” or “10” to a survey question. That is, I am only looking at the percentage of people that selected one of the top two boxes.  I am NOT trying to see if there is statistical significance or calculate the margin of error in the change in Net Promoter Score over time.

The concepts I explore are not just for survey data; I just happen to have some good longitudinal survey data that is well-suited for seeing how to build a stat test formula in Tableau.

I hope you will indulge me and accept that “the company stat guru” has a fine reason for applying a particular statistical test to the data we’ll be analyzing. That said, you should push back on “business-as-usual” assumptions to determine if what you are visualizing and testing really is important (this is the focus of the work Stacy Barr is doing with her Measure Up blog and is the foundation for Stephen Few’s most recent book Signal.)

So, with the assumption that the particular stat test we want to apply – or any stat test, for that matter – is warranted, how do you show it and how do you build it?

Let’s first explore the working dashboard then see how to build it with Tableau.

Note: A very heartfelt thanks to Kelly Martin,, Joe Mako, Vicki Reinhard, Susan Ferarri, and Tiffany Spaulding who helped vet the dashboard.  I went through many different approaches before settling on the one shown below.

A very special thanks to Jeffrey Shaffer who reviewed the blog post and asked some very good questions, and also to Helen Lindsay who provided sample data.

## The data and what we want to show

The data below contains the first few rows of Net Promoter Score survey data with fields for date and role.

Figure 1 — Net Promoter Score survey data with dates and roles

For the dashboard I built I only focused on the percentage of people that were Promoters; that is, people who responded with a 9 or 10 when asked if they would recommend a product or service.

I decided to look at the data broken down by quarters as this particular data set didn’t lend itself to month over month comparison.  Note that the techniques we’ll see will work for any time period.

Here’s the top portion of the interactive dashboard.

Figure 2 — Top portion of dashboard.  Notice that you can change the selected period, the confidence percentage, and filter by company.

### Understanding the chart

Figure 3 — The key features of the chart

Let’s review what we can glean from the chart.  We can see

• The percentage of promoters for a particular period and sort them by role, using a bar chart.
• Which roles have a percentage of promoters that is greater than the previous period and which have less, using color to distinguish (blue for greater, brown for less).
• Just how much more or less the percentage for this period is compared to the previous using a reference line (the bar is the current period; the vertical line is the previous period).
• Which roles showed a significantly significant increase or decrease (the red dot).

Note that that the chart uses “Cotgreavian” tooltips that allow you to glean more detail for a particular role when you hover over a bar:

Figure 4 — Hover over a bar for in-depth information about the role for the current period and the previous period

So, we can see from the red dot that something is up with Lawyers, Doctors and Nurses; that is, the percent increase from the previous period for Doctors and Lawyers is statistically significant and the percent decrease for Nurses is also significantly significant.  Is this a one-time thing or a trend?

### Looking at changes over time

Clicking a role or roles will display trends for that role / roles.  For example, if we select Nurse in the top chart a second chart showing percentage of promoters over time will appear, as shown here.

Figure 5 — Percentage of nurses that are promoters, over time.

The big takeaway for me is that up until the first quarter of 2013 there were very few responses and after that there was both a consistent number of responses along with a consist decline in the percentage of nurses that were promoters.

Should you be hyperventilating because of the four-month downward trend?  That discussion is beyond this blog post but I again encourage you to check out the work Stacy Barr is doing at her Measure Up blog as well as Stephen Few’s most recent book Signal.

## How the This Quarter vs. That Quarter Chart is Built

Let’s dig into how to build this in Tableau, starting with the top viz in the dashboard.

Figure 6 — What’s under the hood.

1. Promoters – Current Quarter. This is the measure that drives the bars.  It’s also driving what appears on the labels.
2. Promoters – Previous Quarter. This measure is on the Level of Detail and drives the reference lines.
3. Greater / Less. This is a discrete measure that determines the color of the bar.

### Promoters – Current Quarter

What we want is the percentage of people that were promoters for the selected quarter, the “selected” quarter being determined by a parameter that the user can control.

Specially, we want to add up everybody that responded with a 9 or 10 for the selected quarter and divide by the total number of people that responded.  Here’s the calculation that handles this.

SUM(

IF [Value]>=9 and DATETRUNC(‘quarter’, [Select Period])==DATETRUNC(‘quarter’,[Date])
then 1 else 0
END)

/

SUM(

IF DATETRUNC(‘quarter’, [Select Period])==DATETRUNC(‘quarter’,[Date])
then 1 else 0
END)

The translation into English is

Take the sum of

If the value from a respondent is greater than or equal to 9 and the date value, truncated to the nearest quarter from the parameter drop down [Select Period] is the same as the date value, truncated to the nearest quarter for [Date], then 1, else 0.

Divide this by the sum of

If the date value, truncated to the nearest quarter for the selected period is the same as the date value, truncated for the nearest quarter for [Date], then 1, else 0.

Not sure about the [DATETRUNC] function vs. the [DATEPART] function?  Have a look at Joshua Milligan’s excellent post explaining date values vs. date parts.

### Promoters – Previous Quarter

This calculation is very similar to the calculation for the Current Quarter, except we want to find results for the quarter that occurred just prior to the selected quarter.  Here’s the calculation.

SUM(

IF [Value]>=9 and DATETRUNC(‘quarter’, [Select Period])=DATETRUNC(‘quarter’,DATEADD(‘quarter’,1,[Date]))
then 1 else 0
END)

/

SUM(

then 1 else 0
END)

The formula is the same except we use the DATEADD function to add an additional quarter; that is, we’re saying that we only want to find results where, when we add an additional quarter, we get a value equal to the current quarter; i.e., the previous quarter, plus one quarter, gives us the current quarter.

### Greater / Less

The color of the bars is determined by this discrete measure:

IF [Promoters — Current Quarter] > [Promoters — Previous Quarter] then “Greater than previous”
else “Less than previous”
END

Yes, I suppose we should have a contingency for when the percentage of promoters for the current period is the same as the previous period; I leave it as an exercise for the reader to add this functionality.

So, we’ve explained everything except … The Red Dot.

## The Red Dot – Computing Statistical Significance on the Fly

Most of my clients and students are surprised to find out that you can fashion a test for statistical significance inside Tableau and it can test for statistical significance “on the fly”; e.g., you can apply filters and Tableau will recalculate based on the filter settings.

The first step is determining just how the client wants to test for statistical significance. This usually entails sending an inquiry to “the stats person” who responds with something that looks like this:

Figure 7 — Z-test formula for statistical significance

I hope your eyes aren’t glassing over as this really isn’t very complicated; it just might look complicated if you’re not used to seeing stat formulas with square root symbols.  Here are the critical things you need to know:

p1            Percentage of promoters for the current period

p2            Percentage of promoters for the previous period

n1            Number of respondents for the current period

n2            Number of respondents for the previous period

If z1 is greater than or equal to 1.96 then there is a 95% degree of confidence that the difference between the two periods is statistically significant.

So, how do we build this formula?

Slowly, and in easy-to-digest pieces.

### The Dot Itself

Figure 8 — The discrete measure Z-Test Significance Dot is responsible for displaying the dot

The calculation that produces the dot is called Z-Test Significance Dot and it is defined as follows.

IF ABS([Promoters — Z-Score Quarter])>=[Confidence] THEN “•”
ELSE “”
END

This translates as

If the absolute value of [Promoters – Z-Score Quarter] is greater than or equal to the confidence parameter (currently set to 1.96, or 95%) then display a dot; otherwise, display a null string.

And just how is [Promoters – Z-Score Quarter] defined?  Let’s explore the next layer of the onion.

### Promoters – Z-Score Quarter

This is defined as follows:

[Promoters — Z-Score Quarter Numerator] /

SQRT(

([Promoters — Z-Score Quarter Denom – Current] +
[Promoters — Z-Score Quarter Denom – Previous])
)

Here’s how it maps to the stat formula we saw earlier:

Figure 9 — Mapping the components of the formula to different calculated field

So now we just need to understand the three different pieces that go into the stat function.

### Promoters – Z-Score Quarter Numerator

This is very simple and refers to calculations we’ve already used.

[Promoters — Current Quarter] –
[Promoters — Previous Quarter]

### Promoters — Z-Score Quarter Denom – Current

This is fairly straightforward given what we’ve already explored.

([Promoters — Current Quarter]*(1-[Promoters — Current Quarter]))
/SUM([Promoters — Current Quarter Count])

Where [Promoters – Current Quarter Count] is defined as follows.

IF DATETRUNC(‘quarter’, [Select Period])==DATETRUNC(‘quarter’,[Date])
THEN 1 END

So SUM(Promoters — Current Quarter Count]) is just adding up all the people that responded during the selected quarter.

### Promoters — Z-Score Quarter Denom – Previous

([Promoters — Previous Quarter]*(1-[Promoters — Previous Quarter]))/
SUM([Promoters — Previous Quarter Count])

This uses the same logic as [Promoters – Z-Score Quarter Denom – Current] but instead aggregates results from the previous quarter.

## Putting it all together

In addition to building the components in a piecemeal fashion I will often build a crosstab of all these components to see if they are working as I would expect.  Consider the crosstab shown here.

Figure 10 — Crosstab showing all the pieces that contribute to the red dot

The cross tab allows us to examine all the intermediate calculations to see how the contribute to the determining calculation in the last column.

## What about the secondary chart?

So we’ve now seen how to build the top chart that shows current and previous quarters broken down by role.  How does the secondary chart – the chart that appears when you click a role or roles in the first chart – work?

Figure 11 — Percentage of promoters for Nurses over time

Here we have a dual axis chart so that we can have both a line (gray) and a circle (colored based on whether the change for the previous period is statistically significant).

In this case we have to construct all of the pieces using a table calculation, but the process of putting together the different components is identical to what we saw earlier.  For example, the calculation that determined the color of the circle, [LONG_Z-Test Significance], is defined as follows.

IF ABS([LONG_Z-Score])>=[Confidence] then “Significant”
else “Not significant”
end

And [LONG_Z-Score] is defined this way:

[LONG_Z-ScoreNumerator] /

SQRT(

([LONG_Z-Score Denom Current] +
[LONG_Z-Score Denom Previous])

)

I also built a crosstab to see how all the pieces fit together, as shown below.

Figure 12 — Crosstab to help put together a z-test calculation for values shown over time

## Conclusion

The dashboard in this blog post shows the percentage of promoters, sorted by role, for a particular quarter, compared with the percentage of promoters for the previous quarter.  Roles where the percentage difference is statistically significant are marked with a red dot. You can drill down on a particular role (or role) and see how scores have changed over time.

While the critical visual component was showing bars and reference lines, most of the “heavy lifting” went into determining if a change was statistically significant.  The key here was to not be intimidated by a statistical formula and to build the calculations in small pieces, using crosstabs to check the work.

Tagged with: , , , ,

“With great power comes great responsibility”

— Voltaire

— Benjamin Parker (Uncle Ben from Spiderman)

## Overview

Recently both Ryan Sleeper and Andy Kriebel blogged about donut charts in Tableau.

Figure 1 — Donut chart courtesy of Andy Kriebel

While both of them cautioned about where, when, and how best to use them, I fear many people will ignore the warnings and dig into this sugary, analytically-impoverished chart type and start creating stuff like this.

Figure 2 — Really bad donut chart. In fact, it’s just a pie chart with a hole in the middle.

Yuk.

And what fuels my fear?  Ryan and Andy do great work, and they write great blogs.  They rightfully have a lot of influence in the Tableau community.

But with great power — and influence —  comes great responsibility and I suspect that some people will see Ryan and Andy’s work, ignore their recommendations, and apply the following bit of “logic”:

Ryan Sleeper is a Tableau Iron Viz champion and really cool — and he makes donut charts.

Andy Kriebel is a Tableau Zen master and really cool — and he, too, makes donut charts.

I want to make cool vizzes and be really cool; therefore, I should make donut charts.

[Insert face palm here]

Interviewer: So, what do you have against donut charts?  Don’t you think they look cool?

Me: My problem is that donut charts don’t tell you very much.

Interviewer: Yes, but they look cool!

Me [yelling]: You know what else looks look cool?  Pictures from the Hubble telescope.  Vintage electric basses.  Three-dimensional pie charts! Should I festoon my dashboards with these images, just because they look cool?

Interviewer: Fine, explain to me why this chart types doesn’t work, but I’d like to see an alternative that isn’t BOR-ING!

Me:  Okay, allow me to do the following:

• Explain why donut charts don’t tell you much (or not as much as a bar chart)
• Present a better alternative
• Show how to have your cake (not your donut) and eat it, too

## Why donut charts don’t tell you much

Consider the chart in Figure 1, above.

I always recommend that people ask the following questions when coming up with a visualization:

• Do I need different colors?
• Do I need a legend?
• Do I need measure labels?

Let’s see what happens when we remove the measure labels:

Figure 3 — donut chart without measure labels.

The chart does pass some of the “can I figure this out test”.  For example, it’s easy for me to see that West is around one quarter of the way to goal and that East is a little more than half way.  Where the chart fails is with comparison among regions.  For example, can you tell how much closer North is to its goal than West?  This comparison is particularly hard to determine as it’s very difficult to gauge how much longer one arc is than another arc.

## A better alternative

I think a bar chart with a goal line is easier to grok.  It tells me more and takes up less screen real estate, too.

Figure 4 — Bar chart with goal line.

There’s an added advantage in that I can easily see both the progress towards a goal and that the goal is \$100,000.

Better yet, suppose the goals were different for each region?  Right now they each have a shared goal of \$100,000 but suppose the goal for North is \$125,000 and the goal for East is \$75,000?  With the donut chart, how will you show the actual goal and the progress towards the goal at the same time?

Why is it easy to compare progress across regions using the bar chart?  I’ve discussed this in length here, but the bottom line is that humans are much better at judging the length of bars than they are judging the area of circles or the lengths of arcs.

But does the chart pass the “no measure labels” test?  Have a look.

Figure 5 — Bar chart without measure labels.

While I prefer having labels, it’s pretty easy for me to the following:

• North is more than twice as long as West
• East is a little more than half way
• South is more than a third of the way to goal
• East is about twice as long as West

In other words,  I can draw conclusions more easily from this chart than the donut chart.

## Another Example

Consider the chart below that shows the percentage of confirmed judicial nominees that are women, broken down by president.

Figure 6 – Donut Chart showing Female Judicial Nominees (source: Alliance for Justice)

There are some good stories in here but they are buried.  Compare this with a bar chart that contrasts the different presidents and underscores the differences between Republicans and Democrats.

Figure 7 — Bar Chart showing Female Judicial Nominees (source: Alliance for Justice)

I think this is a lot clearer.

But it is, well, boring.

## Have your cake and eat it, too

I admit that most of my practice has me building stuff that looks more like it would appear in The Economist than in USA Today, but I do understand that you may need to create something that is eye catching.

And I agree that the donut chart is eye catching, but I hate to sacrifice information for the sake of decoration.

Is there a way to get both?

I think there is.  Let’s work on the first example where we were examining progress towards a goal broken down by region.

### Want some sugar?  Try a lollipop chart

Figure 8 — Lollipop chart

Creating a lollipop chart is easy in Tableau. You create a dual axis chart where both measures are identical but you have a different chart type (in this case a bar chart combined with a circle chart).

Figure 9 – Tableau settings for a lollipop chart

##### Try some fun shapes

We can also take the lollipop chart and dress it up with a custom shape, like the one shown below.

Figure 10 — Combination bar and shape chart

While I prefer the lollipops to the runner, I have no problem with the chart shown above because I don’t have to work hard to see both the distance from the goal and to compare among regions.  That is, I did not fight Tableau’s suggested default chart type but instead took it and dressed it up a bit.

## Conclusion

Even if you are tasked with having to create visualizations for mass public consumption I urge your to use caution before creating a donut chart. I understand that you may need something that is more visually arresting than a simple bar chart, but take that as a challenge: find a way to make something that looks cool but does not sacrifice one bit of analytical clarity.

And if you do create a donut chart, please look carefully at what Ryan and Andy did (and did not do) in fashioning theirs.

Tagged with: , , , ,

In Part One of this series I discussed why the Tableau support community is unique and why you should care. In Part Two I shared my thoughts on the early years of the community and how one person in particular set the tone for sharing knowledge and expertise.  In this final post I make recommendations on the things you can do to ensure that the community continues to thrive.

## What you can, and should do, to ensure the community thrives

I rely on this community to inspire me, cheer me on, and help me when I need it.

I don’t want to lose this invaluable asset, so I’m going to enlist you to contribute to its wellbeing, assuming you are not already doing so. Here are some things you can do.

If you can’t find what you need through a web search, ask for help as it will help the community as a whole. While counterintuitive, asking for help will generate a discussion that will lead to solutions that will help not just you but others that are having or will have the same problem you have.

And just where should you ask for help?  Tableau’s community forum is a great place to start.  If you look you will see a lot of Zen Masters who have posted questions, not just answers, through the years.

In addition to asking, if you want to observe noteworthy Tableau activity, make sure to check out the Twitter hashtag #tableau and also check out the list of Tableau-related tweeters Andy Cotgreave has assembled here.

#### Show the love

If someone has helped you or something has inspired you, send them a “thank you” e-mail, launch a tweet, comment that person’s blog, but above all please let the person who helped you know you appreciate what he / she has done (and in my case feel free to send dark chocolate and / or red wine).

Figure 1 — Beers were free at The 2014 Tableau Conference. But I appreciate the sentiment.

Cheering each other on is a big reason the community thrives.

#### Post your work to Tableau Public.

If you create something worthwhile, share it with the world.  Tableau Public makes it easy, and it’s free.

If you recall from Part One, I stated that Tableau Public is a masterstroke in fostering community and visualization excellence in that it provides a free service for people to post their work.  The public will in turn remark on the work, but the really amazing thing is that a Tableau user can download packaged workbooks to see how they work.

Consider this great “how-to” example from Josh Milligan.

Figure 2 — A great “how to” example from Josh Milligan that anybody can download and dissect.

Notice the “Download” button in the bottom right corner.  With Tableau Public I can do more than just interact with the viz; I can download the workbook and see how the person built it.

#### Help others whenever and wherever you can.

You may not be able to pay back the person or people that helped you, but you can help others.  Do not feel pressure to change the world or have the same impact as a Joe Mako or a Jonathan Drummey, but there’s a lot you can do including participating on the Tableau forum, writing a blog, attending a user group meeting (live or virtual), helping a non-profit understand their data, or just commenting on someone else’s work.

With respect to the forum, try “lurking” (just hanging out and observing the various conversations) to see if this might be an outlet for your abilities. If nothing else you’ll learn a great deal.

With respect to blogging, the barrier to entry has never been lower and this is a great way to find your voice and contribute to the community.  Indeed, Andy Cotgreave maintains that if you can have a Google account you can create a blog and publish a post in fewer than three minutes.

Figure 3 — Dan Montgomery, Paul Banoub, Anya A’Hearn, and Jewel Loree stopping traffic and evangelizing blogging at the 2014 Tableau Conference.

#### Do not celebrate or reward mediocre work.

We should, as a community, be working to improve the art and should not reward stuff that isn’t good.  I’m not saying that you should be a jerk (remember, there are no jerks in this community, at least not yet) but if you see something that you know can be better, please let the person – and the world – know what you would do to make it better.

Incidentally, people critique my work all the time and I’m grateful for the feedback.  Indeed, if I have any “high-stakes” dashboards I want to publish I will always ask both colleagues and laypeople to review the work before it goes live (please see the “Usability” section of Your Tableau Public Viz is Ugly *and* Confusing.)

#### Don’t be too hard on yourself

I remember something Joe Mako told me several years ago:

I like Tableau because it allows me to fail faster.

Do not be afraid to fail, and to fail easily and often. It takes time, study, and practice to get good at data visualization and Tableau.  Do not be afraid to post something on Tableau Public and ask for help or criticism.  Most people will offer constructive help and you’ll get better, fast.