Bubble charts, ratios and proportionality

A recent article in the Wall Street Journal about a challenger to the dominant weedkiller, Roundup, contains a nice selection of graphics. (Dicamba is the up-and-comer.)

The change in usage of three brands of weedkillers is rendered as a small-multiples of choropleth maps. This graphic displays geographical and time changes simultaneously.

The staircase chart shows weeds have become resistant to Roundup over time. This is considered a weakness in the Roundup business.

***

In this post, my focus is on the chart at the bottom, which shows complaints about Dicamba by state in 2019. This is a bubble chart, with the bubbles sorted along the horizontal axis by the acreage of farmland by state.

Below left is a more standard version of such a chart, in which the bubbles are allowed to overlap. (I only included the bubbles that were labeled in the original chart).

The WSJ’s twist is to use the vertical spacing to avoid overlapping bubbles. The vertical axis serves a design perogative and does not encode data.  

I’m going to stick with the more traditional overlapping bubbles here – I’m getting to a different matter.

***

The question being addressed by this chart is: which states have the most serious Dicamba problem, as revealed by the frequency of complaints? The designer recognizes that the amount of farmland matters. One should expect the more acres, the more complaints.

Let's consider computing directly the number of complaints per million acres.

The resulting chart (shown below right) – while retaining the design – gives a wholly different feeling. Arkansas now owns the largest bubble even though it has the least acreage among the included states. The huge Illinois bubble is still large but is no longer a loner.

Now return to the original design for a moment (the chart on the left). In theory, this should work in the following manner: if complaints grow purely as a function of acreage, then the bubbles should grow proportionally from left to right. The trouble is that proportional areas are not as easily detected as proportional lengths.

The pair of charts below depict made-up data in which all states have 30 complaints for each million acres of farmland. It’s not intuitive that the bubbles on the left chart are growing proportionally.

Now if you look at the right chart, which shows the relative metric of complaints per million acres, it’s impossible not to notice that all bubbles are the same size.

All these charts lament the high prices charged by U.S. hospitals

A former student asked me about this chart from the New York Times that highlights much higher prices of hospital procedures in the U.S. relative to a comparison group of seven countries.

The dot plot is clearly thought through. It is not a default chart that pops out of software.

Based on its design, we surmise that the designer has the following intentions:

1. The names of the medical procedures are printed to be read, thus the long text is placed horizontally.
   
   
2. The actual price is not as important as the relative price, expressed as an index with the U.S. price at 100%. These reference values are printed in glaring red, unignorable.
   
   
3. Notwithstanding the above point, the actual price is still of secondary importance, and the values are provided as a supplement to the row labels. Getting to the actual prices in the comparison countries requires further effort, and a calculator.
   
   
4. The primary comparison is between the U.S. and the rest of the world (or the group of seven countries included). It is less important to distinguish specific countries in the comparison group, and thus the non-U.S. dots are given pastels that take some effort to differentiate.
   
   
5. Probably due to reader feedback, the font size is subject to a minimum so that some labels are split into two lines to prevent the text from dominating the plotting region.

***

In the Trifecta Checkup view of the world, there is no single best design. The best design depends on the intended message and what’s in the available data.

To illustate this, I will present a few variants of the above design, and discuss how these alternative designs reflect the designer's intentions.

Note that in all my charts, I expressed the relative price in terms of discounts, which is the mirror image of premiums. Instead of saying Country A's price is 80% of the U.S. price, I prefer to say Country A's price is a 20% saving (or discount) off the U.S. price.

First up is the following chart that emphasizes countries instead of hospital procedures:

This chart encourages readers to draw conclusions such as "Hospital prices are 60-80 percent cheaper in Holland relative to the U.S." But it is more taxing to compare the cost of a specific procedure across countries.

The indexing strategy already creates a barrier to understanding relative costs of a specific procedure. For example, the value for angioplasty in Australia is about 55% and in Switzerland, about 75%. The difference 75%-55% is meaningless because both numbers are relative savings from the U.S. baseline. Comparing Australia and Switzerland requires a ratio (0.75/0.55 = 1.36): Australia's prices are 36% above Swiss prices, or alternatively, Swiss prices are a 64% 26% discount off Australia's prices.

The following design takes it even further, excluding details of individual procedures:

For some readers, less is more. It’s even easier to get a rough estimate of how much cheaper prices are in the comparison countries, for now, except for two “outliers”, the chart does not display individual values.

The widths of these bars reveal that in some countries, the amount of savings depends on the specific procedures.

The bar design releases the designer from a horizontal orientation. The country labels are shorter and can be placed at the bottom in a vertical design:

It's not that one design is obviously superior to the others. Each version does some things better. A good designer recognizes the strengths and weaknesses of each design, and selects one to fulfil his/her intentions.

 

P.S. [1/3/20] Corrected a computation, explained in Ken's comment.

Revisiting global car sales

We looked at the following chart in the previous blog. The data concern the growth rates of car sales in different regions of the world over time.

Here is a different visualization of the same data.

Well, it's not quite the same data. I divided the global average growth rate by four to yield an approximation of the true global average. (The reason for this is explained in the other day's post.)

The chart emphasizes how each region was helping or hurting the global growth. It also features the trend in growth within each region.

 

This Excel chart looks standard but gets everything wrong

The following CNBC chart (link) shows the trend of global car sales by region (or so we think).

This type of chart is quite common in finance/business circles, and has the fingerprint of Excel. After examining it, I nominate it for the Hall of Shame.

***

The chart has three major components vying for our attention: (1) the stacked columns, (2) the yellow line, and (3) the big red dashed arrow.

The easiest to interpret is the yellow line, which is labeled "Total" in the legend. It displays the annual growth rate of car sales around the globe. The data consist of annual percentage changes in car sales, so the slope of the yellow line represents a change of change, which is not particularly useful.

The big red arrow is making the point that the projected decline in global car sales in 2019 will return the world to the slowdown of 2008-9 after almost a decade of growth.

The stacked columns appear to provide a breakdown of the global growth rate by region. Looked at carefully, you'll soon learn that the visual form has hopelessly mangled the data.

What is the growth rate for Chinese car sales in 2006? Is it 2.5%, the top edge of China's part of the column? Between 1.5% and 2.5%, the extant of China's section? The answer is neither. Because of the stacking, China's growth rate is actually the height of the relevant section, that is to say, 1 percent. So the labels on the vertical axis are not directly useful to learning regional growth rates for most sections of the chart.

Can we read the vertical axis as global growth rate? That's not proper either. The different markets are not equal in size so growth rates cannot be aggregated by simple summing - they must be weighted by relative size.

The negative growth rates present another problem. Even if we agree to sum growth rates ignoring relative market sizes, we still can't get directly to the global growth rate. We would have to take the total of the positive rates and subtract the total of the negative rates.  

***

At this point, you may begin to question everything you thought you knew about this chart. Remember the yellow line, which we thought measures the global growth rate. Take a look at the 2006 column again.

The global growth rate is depicted as 2 percent. And yet every region experienced growth rates below 2 percent! No matter how you aggregate the regions, it's not possible for the world average to be larger than the value of each region.

For 2006, the regional growth rates are: China, 1%; Rest of the World, 1%; Western Europe, 0.1%; United States, -0.25%. A simple sum of those four rates yields 2%, which is shown on the yellow line.

But this number must be divided by four. If we give the four regions equal weight, each is worth a quarter of the total. So the overall average is the sum of each growth rate weighted by 1/4, which is 0.5%. [In reality, the weights of each region should be scaled to reflect its market size.]

***

tldr; The stacked column chart with a line overlay not only fails to communicate the contents of the car sales data but it also leads to misinterpretation.

I discussed several serious problems of this chart form: 

• stacking the columns make it hard to learn the regional data
  
  
• the trend by region takes a super effort to decipher
  
  
• column stacking promotes reading meaning into the height of the column but the total height is meaningless (because of the negative section) while the net height (positive minus negative) also misleads due to presumptive equal weighting
  
  
• the yellow line shows the sum of the regional data, which is four times the global growth rate that it purports to represent

 

***

PS. [12/4/2019: New post up with a different visualization.]

How to read this cost-benefit chart, and why it is so confusing

Long-time reader Antonio R. found today's chart hard to follow, and he isn't alone. It took two of us multiple emails and some Web searching before we think we "got it".

 

Antonio first encountered the chart in a book review (link) of Hal Harvey et. al, Designing Climate Solutions. It addresses the general topic of costs and benefits of various programs to abate CO2 emissions. The reviewer praised the "wealth of graphics [in the book] which present complex information in visually effective formats." He presented the above chart as evidence, and described its function as:

policy-makers can focus on the areas which make the most difference in emissions, while also being mindful of the cost issues that can be so important in getting political buy-in.

(This description is much more informative than the original chart title, which states "The policy cost curve shows the cost-effectiveness and emission reduction potential of different policies.")

Spend a little time with the chart now before you read the discussion below.

Warning: this is a long read but well worth it.

 

***

 

If your experience is anything like ours, scraps of information flew at you from different parts of the chart, and you had a hard time piecing together a story.

What are the reasons why this data graphic is so confusing?

Everyone recognizes that this is a column chart. For a column chart, we interpret the heights of the columns so we look first at the vertical axis. The axis title informs us that the height represents "cost effectiveness" measured in dollars per million metric tons of CO2. In a cost-benefit sense, that appears to mean the cost to society of obtaining the benefit of reducing CO2 by a given amount.

That's how far I went before hitting the first roadblock.

For environmental policies, opponents frequently object to the high price of implementation. For example, we can't have higher fuel efficiency in cars because it would raise the price of gasoline too much. Asking about cost-effectiveness makes sense: a cost-benefit trade-off analysis encapsulates the something-for-something principle. What doesn't follow is that the vertical scale sinks far into the negative. The chart depicts the majority of the emissions abatement programs as having negative cost effectiveness.

What does it mean to be negatively cost-effective? Does it mean society saves money (makes a profit) while also reducing CO2 emissions? Wouldn't those policies - more than half of the programs shown - be slam dunks? Who can object to programs that improve the environment at no cost?

I tabled that thought, and proceeded to the horizontal axis.

I noticed that this isn't a standard column chart, in which the width of the columns is fixed and uneventful. Here, the widths of the columns are varying.

***

In the meantime, my eyes are distracted by the constellation of text labels. The viewing area of this column chart is occupied - at least 50% - by text. These labels tell me that each column represents a program to reduce CO2 emissions.

The dominance of text labels is a feature of this design. For a conventional column chart, the labels are situated below each column. Since the width does not usually carry any data, we tend to keep the columns narrow - Tufte, ever the minimalist, has even advocated reducing columns to vertical lines. That leaves insufficient room for long labels. Have you noticed that government programs hold long titles? It's tough to capture even the outline of a program with fewer than three big words, e.g. "Renewable Portfolio Standard" (what?).

The design solution here is to let the column labels run horizontally. So the graphical element for each program is a vertical column coupled with a horizontal label that invades the territories of the next few programs. Like this:

The horror of this design constraint is fully realized in the following chart, a similar design produced for the state of Oregon (lifted from the Plan Washington webpage listed as a resource below):

In a re-design, horizontal labeling should be a priority.

 

***

Realizing that I've been distracted by the text labels, back to the horizontal axis I went.

This is where I encountered the next roadblock.

The axis title says "Average Annual Emissions Abatement" measured in millions metric tons. The unit matches the second part of the vertical scale, which is comforting. But how does one reconcile the widths of columns with a continuous scale? I was expecting each program to have a projected annual abatement benefit, and those would fall as dots on a line, like this:

Instead, we have line segments sitting on a line, like this:

Think of these bars as the bottom edges of the columns. These line segments can be better compared to each other if structured as a bar chart:

Instead, the design arranges these lines end-to-end.

To unravel this mystery, we go back to the objective of the chart, as announced by the book reviewer. Here it is again:

policy-makers can focus on the areas which make the most difference in emissions, while also being mindful of the cost issues that can be so important in getting political buy-in.

The primary goal of the chart is a decision-making tool for policy-makers who are evaluating programs. Each program has a cost and also a benefit. The cost is shown on the vertical axis and the benefit is shown on the horizontal. The decision-maker will select some subset of these programs based on the cost-benefit analysis. That subset of programs will have a projected total expected benefit (CO2 abatement) and a projected total cost.

By stacking the line segments end to end on top of the horizontal axis, the chart designer elevates the task of computing the total benefits of a subset of programs, relative to the task of learning the benefits of any individual program. Thus, the horizontal axis is better labeled "Cumulative annual emissions abatement".

 

Look at that axis again. Imagine you are required to learn the specific benefit of program titled "Fuel Economy Standards: Cars & SUVs".  

This is impossible to do without pulling out a ruler and a calculator. What the axis labels do tell us is that if all the programs to the left of Fuel Economy Standards: Cars & SUVs were adopted, the cumulative benefits would be 285 million metric tons of CO2 per year. And if Fuel Economy Standards: Cars & SUVs were also implemented, the cumulative benefits would rise to 375 million metric tons.

***

At long last, we have arrived at a reasonable interpretation of the cost-benefit chart.

Policy-makers are considering throwing their support behind specific programs aimed at abating CO2 emissions. Different organizations have come up with different ways to achieve this goal. This goal may even have specific benchmarks; the government may have committed to an international agreement, for example, to reduce emissions by some set amount by 2030. Each candidate abatement program is evaluated on both cost and benefit dimensions. Benefit is given by the amount of CO2 abated. Cost is measured as a "marginal cost," the amount of dollars required to achieve each million metric ton of abatement.

This "marginal abatement cost curve" aids the decision-making. It lines up the programs from the most cost-effective to the least cost-effective. The decision-maker is presumed to prefer a more cost-effective program than a less cost-effective program. The chart answers the following question: for any given subset of programs (so long as we select them left to right contiguously), we can read off the cumulative amount of CO2 abated.

***

There are still more limitations of the chart design.

• We can't directly read off the cumulative cost of the selected subset of programs because the vertical axis is not cumulative. The cumulative cost turns out to be the total area of all the columns that correspond to the selected programs. (Area is height x width, which is cost per benefit multiplied by benefit, which leaves us with the cost.) Unfortunately, it takes rulers and calculators to compute this total area.
  
  
• We have presumed that policy-makers will make the Go-No-go decision based on cost effectiveness alone. This point of view has already been contradicted. Remember the mystery around negatively cost-effective programs - their existence shows that some programs are stalled even when they reduce emissions in addition to making money!
  
  
• Since many, if not most, programs have negative cost-effectiveness (by the way they measured it), I'd flip the metric over and call it profitability (or return on investment). Doing so removes another barrier to our understanding. With the current cost-effectiveness metric, policy-makers are selecting the "negative" programs before the "positive" programs. It makes more sense to select the "positive" programs before the "negative" ones!

***

In a Trifecta Checkup (guide), I rate this chart Type V. The chart has a great purpose, and the design reveals a keen sense of the decision-making process. It's not a data dump for sure. In addition, an impressive amount of data gathering and analysis - and synthesis - went into preparing the two data series required to construct the chart. (Sure, for something so subjective and speculative, the analysis methodology will inevitably be challenged by wonks.) Those two data series are reasonable measures for the stated purpose of the chart.

The chart form, though, has various shortcomings, as shown here.  

***

In our email exchange, Antonio and I found the Plan Washington website useful. This is where we learned that this chart is called the marginal abatement cost curve.

Also, the consulting firm McKinsey is responsible for popularizing this chart form. They have published this long report that explains even more of the analysis behind constructing this chart, for those who want further details.

Who is a millennial? An example of handling uncertainty

I found this fascinating chart from CNBC, which attempts to nail down the definition of a millennial.

It turns out everyone defines "millennials" differently. They found 23 different definitions. Some media outlets apply different definitions in different items.

I appreciate this effort a lot. The design is thoughtful. In making this chart, the designer added the following guides:

• The text draws attention to the definition with the shortest range of birth years, and the one with the largest range.
• The dashed gray gridlines help with reading the endpoints of each bar.
• The yellow band illustrates the so-called average range. It appears that this average range is formed by taking the average of the beginning years and the average of the ending years. This indicates a desire to allow comparisons between each definition and the average range.
• The bars are ordered by the ending birth year (right edge).

The underlying issue is how to display uncertainty. The interest here is not just to feature the "average" definition of a millennial but to show the range of definitions.

***

In making my chart, I apply a different way to find the "average" range. Given any year, say 1990, what is the chance that it is included in any of the definitions? In other words, what proportion of the definitions include that year? In the following chart, the darker the color, the more likely that year is included by the "average" opinion.

I ordered the bars from shortest to the longest so there is no need to annotate them. Based on this analysis, 90 percent (or higher) of the sources list 19651985 to 1993 as part of the range while 70 percent (or higher) list 19611981 to 1996 as part of the range.

 

 

Women workers taken for a loop or four

I was drawn to the following chart in Business Insider because of the calendar metaphor. (The accompanying article is here.)

Sometimes, the calendar helps readers grasp concepts faster but I'm afraid the usage here slows us down.

The underlying data consist of just four numbers: the wage gaps between race and gender in the U.S., considered simply from an aggregate median personal income perspective. The analyst adopts the median annual salary of a white male worker as a baseline. Then, s/he imputes the number of extra days that others must work to attain the same level of income. For example, the median Asian female worker must work 64 extra days (at her daily salary level) to match the white guy's annual pay. Meanwhile, Hispanic female workers must work 324 days extra.

There are a host of reasons why the calendar metaphor backfired.

Firstly, it draws attention to an uncomfortable detail of the analysis - which papers over the fact that weekends or public holidays are counted as workdays. The coloring of the boxes compounds this issue. (And the designer also got confused and slipped up when applying the purple color for Hispanic women.)

Secondly, the calendar focuses on Year 2 while Year 1 lurks in the background - white men have to work to get that income (roughly $46,000 in 2017 according to the Census Bureau).

Thirdly, the calendar view exposes another sore point around the underlying analysis. In reality, the white male workers are continuing to earn wages during Year 2.

The realism of the calendar clashes with the hypothetical nature of the analysis.

***

One can just use a bar chart, comparing the number of extra days needed. The calendar design can be considered a set of overlapping bars, wrapped around the shape of a calendar.

The staid bars do not bring to life the extra toil - the message is that these women have to work harder to get the same amount of pay. This led me to a different metaphor - the white men got to the destination in a straight line but the women must go around loops (extra days) before reaching the same endpoint.

While the above is a rough sketch, I made sure that the total length of the lines including the loops roughly matches the total number of days the women needed to work to earn $46,000.

***

The above discussion focuses solely on the V(isual) corner of the Trifecta Checkup, but this data visualization is also interesting from the D(ata) perspective. Statisticians won't like such a simple analysis that ignores, among other things, the different mix of jobs and industries underlying these aggregate pay figures.

Now go to my other post on the sister (book) blog for a discussion of the underlying analysis.

 

 

Powerful photos visualizing housing conditions in Hong Kong

I was going to react to Alberto's post about the New York Times's article about economic inequality in Hong Kong, which is proposed as one origin to explain the current protest movement. I agree that the best graphic in this set is the "photoviz" showing the "coffins" or "cages" that many residents live in, because of the population density. 

Then I searched the archives, and found this old post from 2015 which is the perfect response to it. What's even better, that post was also inspired by Alberto.

The older post featured a wonderful campaign by human rights organization Society for Community Organization that uses photoviz to draw attention to the problem of housing conditions in Hong Kong. They organized a photography exhibit on this theme in 2014. They then updated the exhibit in 2016.

Here is one of the iconic photos by Benny Lam:

I found more coverage of Benny's work here. There is also a book that we can flip on Vimeo.

In 2017, the South China Morning Post (SCMP) published drone footage showing the outside view of the apartment buildings.

***

What's missing is the visual comparison to the luxury condos where the top 1 percent live. For these, one can  visit the real estate sites, such as Sotheby's. Here is their "12 luxury homes for sales" page.

Another comparison: a 1000 sq feet apartment that sits between those extremes. The photo by John Butlin comes from SCMP's Post Magazine's feature on the apartment:

***

Also check out my review of Alberto's fantastic, recent book, How Charts Lie.

 

 

What is a bad chart?

In the recent issue of Madolyn Smith’s Conversations with Data newsletter hosted by DataJournalism.com, she discusses “bad charts,” featuring submissions from several dataviz bloggers, including myself.

What is a “bad chart”? Based on this collection of curated "bad charts", it is not easy to nail down “bad-ness”. The common theme is the mismatch between the message intended by the designer and the message received by the reader, a classic error of communication. How such mismatch arises depends on the specific example. I am able to divide the “bad charts” into two groups: charts that are misinterpreted, and charts that are misleading.

 

Charts that are misinterpreted

The Causes of Death entry, submitted by Alberto Cairo, is a “well-designed” chart that requires “reading the story where it is inserted and the numerous caveats.” So readers may misinterpret the chart if they do not also partake the story at Our World in Data which runs over 1,500 words not including the appendix.

The map of Canada, submitted by Highsoft, highlights in green the provinces where the majority of residents are members of the First Nations. The “bad” is that readers may incorrectly “infer that a sizable part of the Canadian population is First Nations.”

In these two examples, the graphic is considered adequate and yet the reader fails to glean the message intended by the designer.

 

Charts that are misleading

Two fellow bloggers, Cole Knaflic and Jon Schwabish, offer the advice to start bars at zero (here's my take on this rule). The “bad” is the distortion introduced when encoding the data into the visual elements.

The Color-blindness pictogram, submitted by Severino Ribecca, commits a similar faux pas. To compare the rates among men and women, the pictograms should use the same baseline.

In these examples, readers who correctly read the charts nonetheless leave with the wrong message. (We assume the designer does not intend to distort the data.) The readers misinterpret the data without misinterpreting the graphics.

 

Using the Trifecta Checkup

In the Trifecta Checkup framework, these problems are second-level problems, represented by the green arrows linking up the three corners. (Click here to learn more about using the Trifecta Checkup.)

The visual design of the Causes of Death chart is not under question, and the intended message of the author is clearly articulated in the text. Our concern is that the reader must go outside the graphic to learn the full message. This suggests a problem related to the syncing between the visual design and the message (the QV edge).

By contrast, in the Color Blindness graphic, the data are not under question, nor is the use of pictograms. Our concern is how the data got turned into figurines. This suggests a problem related to the syncing between the data and the visual (the DV edge).

***

When you complain about a misleading chart, or a chart being misinterpreted, what do you really mean? Is it a visual design problem? a data problem? Or is it a syncing problem between two components?

Three estimates, two differences trip up an otherwise good design

Reader Fernando P. was baffled by this chart from the Perception Gap report by More in Common. (link to report)

Overall, this chart is quite good. Its flaws are subtle. There is so much going on, perhaps even the designer found it hard to keep level.

The title is "Democrat's Perception Gap" which actually means the gap between Democrats' perception of Republicans and Republican's self-reported views. We are talking about two estimates of Republican views. Conversely, in Figure 2 (not shown), the "Republican's Perception Gap" describes two estimates of Democrat views.

The gap is visually shown as the gray bar between the red dot and the blue dot. This is labeled perception gap, and its values are printed on the right column, also labeled perception gap.

Perhaps as an after-thought, the designer added the yellow stripes, which is a third estimate of Republican views, this time by Independents. This little addition wreaks havoc. There are now three estimates - and two gaps. There is a new gap, between Independents' perception of Republican views, and Republican's self-reported views. This I-gap is hidden in plain sight. The words "perception gap" obstinately sticks to the D-gap.

***

Here is a slightly modified version of the same chart.

 

The design focuses attention on the two gaps (bars). It also identifies the Republican self-perception as the anchor point from which the gaps are computed.

I have chosen to describe the Republican dot as "self-perception" rather than "actual view," which connotes a form of "truth." Rather than considering the gap as an error of estimation, I like to think of the gap as the difference between two groups of people asked to estimate a common quantity.

Also, one should note that on the last two issues, there is virtual agreement.

***

Aside from the visual, I have doubts about the value of such a study. Only the most divisive issues are being addressed here. Adding a few bipartisan issues would provide controls that can be useful to tease out what is the baseline perception gap.

I wonder whether there is a self-selection in survey response, such that people with extreme views (from each party) will be under-represented. Further, do we believe that all survey respondents will provide truthful answers to sensitive questions that deal with racism, sexism, etc.? For example, if I am a moderate holding racist views, would I really admit to racism in a survey?