Junk Charts

Round up

Here are some interesting reading from other places:

Tag clouds have caught on since we approved them a while ago.  One interesting use was at the Life Vicarious blog.  They use it to compare the inclinations of three New York-based restaurant reviewers.  What they should have done is to remove irrelevant words like  "one", "also", "many", "make"/"made", etc.  In statistics, this is called removing "noise" which helps bring out the "signal".

Andrew Gelman discussed the NYT article that reported the finding of unexpected male bias in the children of Asian American families.  He can be counted on to make useful comments on any accompanying graphics.  He rightly pointed out that this is one example of not starting at zero: the relevant baseline is 100 since the metric is essentially the over-age of males relative to females.  I also agree that a line chart with a longer time series plotting percentages rather than over-age would work better.

The racetrack chart made an appearance at Flowing Data.  This one is even more busy and just as impossible to decipher.

The trouble with two lines

It's never a good idea to put two scales on one chart, and this is another example of what not to do:

The ugliest part of this chart are the duelling gridlines. Because neither axis starts at zero, it is difficult to access whether the number of newspapers was declining at a faster pace than the circulation was. Also, line charts would be better able to trace the evolution over time. The interspersed blue and red columns interfere with each other. Note that the designer lessened our pain by plotting every other year, thus halving the number of columns.

The junkart version also puts two data series on the same chart but on the same scale. Instead of plotting the raw data, we plot indices, with 2008 as 100. This reveals a pattern that was not apparent in the original chart. There appeared to have been four periods of evolution: up till 1980, both the number of newspapers and total circulation were at a plateau; from 1980 to 1990, the circulation stayed stable while the number of papers dropped drastically, indicating perhaps consolidation; then from 1990 to 2003, both series declined at roughly equal rates; and finally, the bottom dropped off the circulation from 2003.

See our previous discussion of dual axes here, here, and here

Reference: "Channel Shift: Online Circulars: the first step by retailers toward web-to-store harmony? ", Internet Retailer (print), May 2009. Data from the Newspapers Association of America.

Bonus 1: Thinking about the column interspersing trick some more, I realize that it is, em, possible, em, that one series was plotted for odd years and the other series plotted for even years!

Bonus 2: Here is the requested scatter plot (still indiced):

Some things never fall

It's New York City.  House prices do not fall.  Enjoy the graphic!

Click here for the interactive version

Reference: "For Housing Crisis, the End Probably Isn't Near", New York Times, April 21 2009.

PS. Memo to grammar teachers: When did they start using contractions in titles? 

Don't mess with the scale

My friend Patrick pointed out the single biggest issue with the chart below -- that the designer chose a scale that precisely undermines the message of the chart.  Undermine may be too mild a word to use here; annihilate may be more apt.

The lines in this chart are anchored at the zero point on the time line (horizontal axis) used to indicate the bottoms of various bear markets in the Dow from 1929 to 2007.  From that anchor, time runs to the left showing the amount of time for the Dow to go from peak to bottom (the decline); time runs to the right showing the amount of time for the Dow to climb back to the prior peak (the recovery).  As the caption said, the point of the chart is "if the decline was fast, the recovery took a considerable time".

Funny thing then that the distances from the zero point are roughly comparable on the left as on the right.

This illusion resulted from some very convoluted and perplexing messing around with the horizontal scale.  First, the left-of-center scale is in months while the right-of-center is in years.  Second, the left-of-center scale has normal spacing while the right-of-center seemingly was suffering from spasms.  Take a closer look:

The first five years (0-5) took up about half the scale while the next five (5-10) took maybe one-eighth.  The first year (0-1) took about as much space as the next two years (1-3).

I am not quite sure what is the logic behind this but since the message of the chart has everything to do with the time duration, it is most unfortunate to introduce such distortions.

There is yet another "innovation" in this chart.  Notice that on the right side, the axis labels are irregular (more spasms)... 0,1,3,4,5,10,15,20, 25...  This is as if the designer is posing one of those IQ questions requiring readers to figure out the next number in the sequence.  The specific time intervals selected may have meaning: note that all the lines are straightened out in between these tick marks.  Given that each line represents a different historical sequence, it is difficult to comprehend the regularity of these intervals across history.  Perhaps this will prove to be the key to unlocking the secret of this chart.  Please comment below if you are able to unravel this mystery.

Besides, the same type of "innovation" was not applied to the left side of the chart.  Here, the designer opted to throw out all the data between the peak and the bottom and straightened out all the intermediate fluctuations.

Below are two different versions of this chart, basically restoring the time scale to the normal, equally spaced, symmetric appearance.  The top one used monthly Dow returns where the volatility obstructed our understanding of the trends, requiring the use of color to differentiate the lines.  In the next version, I used R to generate the loess estimates (a type of smoothing) and the trends became clearer.  (There was a prior discussion of loess on Junk Charts here.)

 

Now, these pictures are very different from the original graph!

I'd be very cautious about reading into these charts anyway.  This question is not one suitable for statistical analysis.  The sample size of six is far too small.  Each recession is different in terms of causes, remedies and context.  The fact that we call them recessions do not make them comparable.  Further, it is also impossible to know at this stage if the 2007 decline has reached bottom.  The chart designer essentially assumed this to be the case but who knows?

PS. Nick Rapp, one of the designers of the chart, responds in the comments.  He has started a blog to feature the work of his graphics team at AP.  His colleague has created an interactive version.  More than anything, this post highlights an aspect of the chart that Nick and his team clearly spent a lot of time doodling over.  The concept of the chart itself is wonderful actually, if I didn't say so already; it is essentially the same chart as the oft-printed chart where the anchor point is the start of each recession, only here the anchor is the bottom of each recession.

Exemplary

From Bernard L., another exemplary effort by the Times.  This one really got me excited.  

The set of line graphs shows how demographics of students in American schools have evolved in the last two decades.  Here, I selected New York City schools, and the tool sensibly decided to compare those with New York State schools (gray line).

There is so much to learn from one simple chart:

• The blue and gray lines are almost parallel everywhere, which tells us that in terms of the change in demographic composition, New York City pretty much resembled New York State during this entire period.
• However, in terms of demographic composition, rather than the change in composition, New York City schools are very different from the rest of the state, in that the proportion of white is lower by a third while that of minorities are much higher, especially black and Hispanic students.
• State-wide (as well as city-wide), black and white students have been declining as a proportion while Hispanics and Asians have increased. 
• The extent of the change is immediately visible, Asians have jumped from 7% to 14% for example.  

  
From a graph design perspective, the execution is very clean.  Data labels are limited to the first and last values.  A small multiples concept is used with the ethnic groups placed side by side.  A great awareness of foreground and background as well.  And imagine how much data has been visualized here, and be impressed.  You can look at any county in the country.

Here's one where the county change does not exactly mirror the state change (Napa in California):

Reference: "Diversity in the classroom", New York Times, March 12 2009.

Deep dive

The chart on the right which compares the unemployment picture across past recessions has many good features.



It uses a very sensible metric, counting the percentage change from peak employment.  I have mentioned the superiority of this type of presentation compared to plotting a time series of unemployment rates.  The following is an example of a standard graphic using gray bands to indicate recessions.  The difference is obvious.

Here is the previous post that dealt with the drop in market capitalization of banks since the peak which screamed out for this type of treatment.

It handles the foreground-background issue very well.  A number of similar charts is circulating in which every single line has a different color.  Here, the designer clearly tells us the current recession is the foreground and all the past recessions form the background.  It looks as if the 1981-3 recession is slightly highlighted with a darker orange to draw attention to the fact that it is most similar to the current situation.  I find this unnecessary because the association is clear even without the darker hue; however, the designer does this with a very light touch so this is just a question of taste.

I would label the horizontal axis starting at 0. 


Reference: "Job Losses Hint at Vast Remaking of Economy", New York Times, Mar 8 2009.

Bond yield makeover

Message to readers: I have a large backlog of reader suggestions.  Please be patient as I slowly get through them.  The frequency of posts will remain lower for the time being as I am busy finalizing a draft of a book.  More on that in the near future.

Matt H, a reader, sent in the following entry (with minor edits).

I saw a couple of bad charts on money.cnn.com and thought I'd submit them to you. 

They're both part of the same feature on investment bargains caused by the recession. 

Here are the links:  

Chart A: Municipal Bonds

Chart B: Corporate Bonds

It seems to me like both charts would have made their points more eloquently by using a much simpler, more common form, like a bar chart. 

In Chart A, cubes are used to display the difference between treasury bond yields and AAA municipal bond yields at the two-year horizon and the ten-year horizon.  The volume of each cube corresponds to the yield for the given type of bond in the given period (I think), which spreads the one dimension being compared (yields) across three dimensions, making the differences look smaller than they really are.

[...] At the two-year horizon, the two yields being compared are 1.16% for Treasury bonds and 3.01% for AAA municipal bonds.  The yield for AAA municipal bonds in this case is more than 2.5 times larger than the yield for Treasury bonds, but the difference doesn't look nearly that big in the chart provided.  [...] 

Time out.  Let me add that the inadvertent reference to an optical illusion concerning foreground and background!  The "outline only" cube on the left should have approximately the same volume as the "solid red" cube on the right (3.01% versus 3.30%)  and yet the red cube appeared quite a bit larger because our eyes reacted to the solid color more than thin outlines.

Matt continued:

In Chart B, [...] Again, the metric in question is bond yields:  ten-year Treasury bond yields compared to investment-grade corporate bond yields.  The 2008 figure for each is shown alongside the five-year average.  This chart uses the area of a circle to express these yields, spreading the one-dimensional value across two dimensions.  As in Chart A, the result is a chart in which the difference between values does not appear as large as it actually is.  

I will also send a simple bar chart version of each chart -- the bar charts should illustrate the differences in yields more effectively than the charts actually used in this article.

These are his revised charts:

We can do even better to convert the chart on the right to a time-series line chart.  Instead of the five-year average, it is better to display the gap beween treasury and corporate bonds for each of the five years plus 2008.  This should make for a more eye-catching graphic.

Reference: "Investment in the bargain bin", CNN Money.

Seen at Starbucks

Merry Christmas!

If you happened to be in a Starbucks recently, you might have picked up some charts, which was what happened to one of our regular readers and commentators, ZBicyclist, who then tried his hand at chart critique here.  He is worried they may reach millions soon.  So look out!

This graph on the right -- which should rightfully be called a "tree ring graph" -- seems to me a fantastic concept although it is hard to think of data that would deserve this treatment.  Certainly not the retail sales series plotted here! 

This is an object lesson in why bubble charts often fail.

One issue is scale: note the awkward way in which the innermost ring is used to designate the oldest sales data of $375 billion presumably in 1996, and think about how you would decide where to place the 2007 ring.  (It's arbitrary.)

Another problem is labeling: when the growth is slow, the rings are close together, and labels have to be jittered (look at 2001 and 2002).  In this case, a relatively simple solution is to have the entire series of years run diagonally.

Yet another challenge is relative radii versus relative areas.  Inevitably, some readers will respond to the areas while others will respond to radii  ZBicyclist, for example, belongs to the first group while in this case, I find myself siding with the latter.  When the bubbles/rings overlap, it is difficult to assess areas.

Of course, a simple line chart would do the job with minimal fuss.  The following chart issued by the National Retail Federation actually plots the growth rates, rather than the annual sales.

(Please lose the grid-lines, or else add a vertical axis and lose the data labels.  A column or dot chart is also slightly preferred.)

Now go read ZBicyclist's point of view.  In the meantime, let us know if you think of ways to use the tree ring graph.

Food art

Adam, who is the designer behind the Wired graphics special on "The Future of Food", asked about the rest of the series.  We previously made some comments on a set of mini donut charts.

The first thought that came to mind after browsing through all the charts was: what a great job they have done to generate interest in food data, which has no right to be entertaining.  Specifically, this is a list of things I appreciated:

• An obvious effort was undertaken to extract the most thought provoking data out of a massive amount of statistics collected by various international agencies.  There weren't any chart that is overstuffed, which is a common problem.
• It would be somewhat inappropriate to use our standard tools to critique these charts.  Clearly, the purpose of the designer was to draw readers into statistics that they might otherwise not care for.   Moreover, the Wired culture has long traded off efficiency for aesthetics, and this showed in a graph such as this, which is basically a line chart with two lines, and a lot of mysterious meaningless ornaments:
• 
  
• A nice use of a dual line chart, though.  It works because both data series share the same scale and only one vertical axis is necessary, which is very subtly annotated here.
• The maintenance of the same motifs across several charts is well done.  (See the pages on corn, beef, catfish)
  

Further suggestions:

• It would be nice if Wired would be brave enough to adopt the self-sufficiency principle, i.e. graphs should not contain a copy of the entire data set being depicted.  Otherwise, a data table would suffice.  The graphical construct should be self-sufficient.  This rule is not often followed because of "loss aversion"; there is the fear that a graph without all the data is like an orphan separated from the parents.  Since, as I noted, these graphs are mostly made for awe, there is really no need to print all the underlying data.  For instance, these "column"-type charts can stand on their own without the data (adding a scale would help).
• Not sure if sorting the categories alphabetically in the column chart is preferred to sorting by size of the category.  The side effect of sorting alphabetically is that it spreads out the long and the short chunks, which simplifies labelling and thus reading.
• Not a fan of area charts (see below).  Although it is labelled properly, it is easy at first glance to focus on the orange line rather than the orange area.  That would be a grave mistake.  The orange line actually plots the total of the two types of fish rearing, not the aquaculture component.  The chart is somewhat misleading because it is difficult to assess the growth rate of aquaculture.  Much better to plot the size of both markets as two lines (either indiced or not).
  
•  

Reference: "The Future of Food", Wired, Oct 20 2008.

Mind the gap

When comparing two time series, one typically wants to discuss the size of the gap as it changes over time.  This Business Week chart, for example, depicted for readers the expanding gap between intra-day high and low prices of the S&P 500 for 2008.


This chart construct is effective at pointing out large changes but lacks precision in conveying smaller differences, or trends.  It is always a good idea to plot the gap directly, as we will show below.

More importantly, a better choice of scale can help a lot.  By focusing exclusively on variability (extreme values), this chart hides the relevant information of the closing prices of the S&P.  A point spread of a 100 points means more when the index is at 800 than at 1200.  In order to capture this, we can divide the point spread by the opening price of that day so we say the gap is one-eighth or one-twelfth of the opening price. 

The junkart version makes both changes.  The top chart fixes the scale, plotting the point spread as a percentage of daily opening prices.  Relative to the original chart, the variability in the front part of 2008 was muted because the index was at higher levels back then. 

The bottom chart plots the gap sizes (lengths of the high-low lines).  It is without doubt that directly plotting the gaps showcases the key message.  The current level of volatility is more than double what occurred at the beginning of the year.

If one wants to illuminate the trend as opposed to daily fluctuations, a further improvement will be using moving averages.

For those interested, shown below is a scatter plot that compares the original point spread and the derived point spread, which shows that the change is not trivial.

 

Reference: "The Market: A Daily Roller Coaster", Business Week, Oct 27 2008.