Why I read the daily newspaper comics

When I was a kid, we called them “the funny pages” whether they were or not. I still read them daily, although I’m starting to skip some, especially when my grandson and I curl up with “the funnies” together. You try explaining Dr. Morgan to a three-year-old–it’s beyond my abilities.

Why bother with the daily comics any more? I know they’re a ridiculous waste of time, but sublime moments like the one below make it all worthwhile. Too bad that Roy Lichtenstein couldn’t see this panel.

Dr-Morgan

Background: This is part of the current storyline about spousal abuse, the less common kind. Despite being a year older than I am, Rex Morgan is still pushing the envelope.

Two axes to grind

Published chart

Fig. 1: “A Chart That Says the War on Drugs
Isn’t Working,” as published, with two y-axes.
(Click each image for a larger view.)

By any measure, the War on Drugs has been and continues to be a failure. Although an ever-increasing number of political figures is willing to admit this, the argument has not been presented to the public as clearly as it might.

Recently, for example, salon.com drew attention to a chart reproduced in theatlanticwire.com.  The graphic (Fig. 1), which I’ve recreated here for easier comparison*, purports to show how the 40-year-old “war” has been, according to salon.com, “an extremely expensive Huge Government boondoggle.” While annual federal spending to “control” drugs has increased 20-fold during that period, the national rate of addiction has remained largely unchanged at between 1% and 2% of the U.S. population.

Alarmist chart

Fig. 2: This chart exaggerates the disparity between the two sets of data and gives the impression of an even greater degree of waste.

Figure 1 shows that relationship, sort of, but it misleads the viewer. How? By distorting the relationship between the two sets of numbers, using different units of measurement. Instead of comparing percentages to percentages or dollars to dollars, the published chart “compares” percentages to dollars, each according to its own arbitrary scale, as shown on the left and right y-axes, respectively.

To see how pernicious a practice this is, consider how manipulating the chart’s two y-axes scales can reinforce contradictory conclusions.

An “Alarmist Version” of the original chart (Fig. 2) heightens the suggestion that the War on Drugs has been tremendously wasteful.

Dismissive chart

Fig. 3: This chart minimizes the relationship between the two sets of data, suggesting ineffectiveness perhaps, but negligible waste.

A “Dismissive Version” (Fig. 3), on the other hand, suggests that the War on Drugs has been hardly wasteful at all.

I’m sure the original chart was designed in good faith. The fact remains, however, that it distorts the data it supposedly illustrates, a inherent flaw of charts with two y-axes scales. What makes the published chart especially disappointing is that it carelessly assumes that its depiction is value-free.

Spending per Addict

Fig. 4: This chart shows a large increase in control dollars spent per addict over a time when the addiction rate remained essentially flat.

What would be a better way of expressing these facts, one without bias (inasmuch as this is possible)?

Consider “U.S. Drug Control Spending per Addict, 1970-2010″ (Fig. 4). By converting the two sets of data to a common unit of measurement, control dollars spent per addict, this chart more accuratetly shows the growth in spending during a time when the rate of addiction remained roughly constant. Not as dramatic as the preceding versions certainly, but considerably more honest.

Can you design a different chart to show this information with minimal bias? What dual y-axes charts have you seen lately?
 
 
 
* Figures 1-4 are based on a close reading of the original theatlanticwire.com chart instead of the actual numbers, which weren’t readily available. Theatlanticwire.com identifies the sources as the U.S. Department of Health and Human Services and the International Centre for Science in Drug Policy by way of documentary filmmaker Matt Groff.

Return

Your graphic interpretation here

This chart recently appeared in the news. Supposedly it shows the destructive potential of a nuclear weapon as calculated by some nameless Iranian scientist.

Turns out it’s a crude copy of what’s called a “normal distribution curve,” commonly found in textbooks and online. In this case, the lines show the relationship of power and energy output over time.

However, the shapes of the two overlapping curves — one brief and limited and the other sustained and open-ended — suggest other associations. What additional relationships might this chart depict?

  • The payoff of sex versus love?
  • The heartfelt significance of a gift card versus a handmade present?
  • The pain of stepping barefoot on a Lego in the dark versus losing an adolescent to independence?

What interpretations can you assign to the chart above?

I have this thing about charts

Statistical graphics are commonly used to “bulk up” news stories and add a sense of scientific authority to their claims. However, most charts in the popular media are junk–hard to read and misleading at the same time. This one (fig. 1), for example, appeared widely after the suicide of former San Diego Chargers linebacker Junior Seau.

The tragedy prompted speculation about how a career’s worth of head injuries caused by collisions against today’s larger opponents might’ve contributed to Seau’s death. But the chart is awful. What are the two average weights supposed to be–286 or 295? 303 or 310?

The first step to improving this chart (after fixing the typo in the title) is to dump the fake 3-D effect (fig. 2). Now you can clearly see that the correct values are 289 and 306.

The 2-D chart is still deceiving, however, because it suggests that the average 2011 player was nearly three times the size of the average 1994 player.

Fixing this is simply a matter of basing the chart’s scale on zero (fig. 3). This makes it obvious that the average 2011 player was roughly 6% bigger than his 1994 counterpart.

This redesigned chart is now accurate and easy to read, which should be the absolute minimum standard for statistical graphics. It’s still a bit dry, but any attempt to “prettify” the numbers should be careful to maintain this standard. Here’s what an accurate, legible, more visually interesting statistical graphic might look like–don’t you agree?

Judged by its cover, architecturally

Santa Maria del Santo Spirito exterior

Few adages have as much applicability as the cliched advice to never judge a book by its cover. Here is the entrance to the Basilica of Santa Maria del Santo Spirito in Florence, based on designs by Filippo Brunelleschi.

Click on the image to see what you find inside the church’s unassuming exterior.