Views > Data visualization > Book review: The Functional Art by Alberto Cairo

Outside the realm of hard sciences, I generally follow a simple heuristic to identify books worth reading.

Read books that have been in print for 15 years or so.
Break rule #1 for books written by authors who have in their portfolio at least one book that passes rule #1.

For example, without hesitation I’d read a book published today by Edward Tufte given that The Visual Display of Quantitative Information has been in print since 1983.

The importance of the heuristic is that it’s enormously economizing—given that time is the most precious commodity. Moreover, the 15-year threshold is guarantee that the books I’m reading are Lindy prone—that is, their ideas will most probably survive the ravages of time.

You may be wondering what am I doing then reviewing a book that is less than four years old. Pure curiosity. I was eager to learn about the author with the guts to criticize Edward Tufte—for the many mistakes he made in his books—and what he has to say about statistical charts, maps and explanation diagrams. And boy was I disappointed.

Edward Tufte once said that “ Graphic competence demands three quite different skills: The substantive, statistical, and artistic.” Therefore, in the following paragraphs I’ll share with you few examples from The Functional Art where Alberto Cairo came short on at least two of the three skills required to produce good graphics.

Here is a typical scenario. Alberto Cairo is reading a book or an article. A given hypotheses is proposed. Alberto—being the journalist characterized by “an insatiable, childish curiosity”—is uncomfortable with the claims made. He decides to put the ideas to the test. For that, he pulls from the web some publically available data sets. Designs a couple of charts. Calculates few ratios. And voilà, Alberto proudly presents the evidence proving or debunking the hypotheses made.

For the uninformed, Alberto Cairo comes across as a confident expert in data analysis. However, when scrutinized, you’ll see that his simplistic, casual and non-rigorous approach to analytical and statistical thinking is very dangerous—especially for journalists, the main target audience of his book, who are generally known for their weak math and statistical skills.

Here are the three examples.

EXAMPLE 1 - BE WARY OF ECOLOGICAL CORRELATIONS

In Chapter 6, Alberto Cairo committed the Ecological Fallacy while testing the validity of the hypotheses made in an article that “Obesity is, on average, inversely proportional to the average education of the population”.

For a detailed analysis of this example click here.

BOOK REVIEW: THE FUNCTIONAL ART BY ALBERTO CAIRO

EXAMPLE 2 – DON’T USE THE WRONG METRICS TO SWAY PUBLIC OPINION
In Chapter 2, Alberto Cairo uses the wrong metrics to reach the wrong conclusions about Brazil’s new strategic defense plan.

For a detailed analysis of this example click here.

EXAMPLE 3 – DON’T AVERAGE AVERAGES

In Chapter 6, Alberto Cairo confuses simple average for weighted average while analyzing the link between obesity and educational attainment in the United States.

I also suspect that Alberto Cairo has committed the same mistake while assessing the premise that the world average fertility rate will converge to the replacement rate of 2.1 child per woman (Chapter 1).

For a detailed analysis of this example click here.

“He uses statistics as a drunken man uses lamp-posts—for support rather than illumination.”

Andrew Lang
Scottish poet, novelist and literary critic

The public trust in journalists is at an all-time low. The proliferation of data journalism is only going to make matters worse. Unless we get serious and rigorous about analytical and statistical thinking, charlatanism and sophistry will be on the rise.