Quetelet's legacy

There's an old quip that there are lies, damned lies, and then there's statistics.

I'm not saying it doesn't have its uses, but the misuse of statistics is a significant problem.  Even how numbers are presented can make a huge difference in how they're perceived -- something that is routinely done to shape public opinion.  Considering the following:

There are three medicines -- A, B, and C -- that are being considered to treat an aggressive form of cancer.  Upon large clinical trials, it is found that over five years following treatment, drug A reduces the risk of recurrence from 94% to 88%, B increases the chances of remaining cancer-free by six percent, and C doubles your chance of staying healthy during that time.

Which one do you choose to take?

It turns out, of course, that the statistics of all three are identical.  Your chance of being cancer-free after taking A is 12%, as compared to 6% without the drug.  That's the same as B -- an increase of 6% in the chance of remaining healthy.  But it's also the same as C, because going from a 6% to a 12% remission rate represents an increase by a factor of two.

But to a lot of people, they all sound different.  Don't fool yourself by thinking this kind of thing isn't being used, deliberately, to mislead.  Especially, any time you see statistics such as, "Doing ____ doubles your risk!", the first thing you should ask is, "What is my risk of the same bad outcome if I don't do _____?"  

After all, twice a very small number is still a very small number.

Things get even more muddled when you throw averages into the mix.  Oh, they have their uses; looking at the average score on a well-constructed test, for example, can tell a teacher if, as a whole, (s)he is teaching the students effectively.  The problem occurs when you start trying to apply averages in situations where they don't belong, such as the statement that the average human has slightly less than one testicle.

As I used to tell my Critical Thinking students, significant > true.

In large part, we owe our incessant focus on turning everything into numbers to two men -- Adolphe Quetelet and Francis Galton.  Quetelet, a Belgian polymath, at least started out with good intentions; he'd noticed how a lot of physical characteristics, from human heights to repeated position measurements of astronomical objects, followed a normal distribution (colloquially called a "bell curve"), where there are a few extreme outliers and a great many values in the middle.  That the ubiquity of this pattern could be due to more than one thing -- in my two examples, that the first was because of the effects of genetics, diet, and body mechanics, and the second due to random measurement error -- he conveniently glossed over.

Adolphe Quetelet (ca. 1870) [Image is in the Public Domain]

Quetelet then took a dangerous leap.  Because this pattern was common, he decided it must be good.  He started measuring everything he could, and found the same pattern showing up in assessments of intelligence, body/mass index, individual wealth, size and position of facial features, and skull shape.  He began an obsessive quest for l'homme moyen -- the "average man," whose characteristics showed the least possible deviation from the norm.

Which Quetelet decided also had to be the "best possible man."

Then Francis Galton took hold of this idea, and ran right off the cliff with it.  Galton was an English statistician and psychologist (and, incidentally, Charles Darwin's cousin), and also a raging racist, who decided to use Quetelet's methods to prove his thesis that other races, especially Black Africans, were inherently inferior to White Europeans.  He wasn't subtle about it.  "The average intellectual standard of the Negro race is some two grades below our own," Galton wrote.  "It is mere heredity.... [Black Africans] are lazy, palavering savages...  It would be for the best if some means could be contrived for the coast of Africa be given to Chinese colonists so that they might supplant the inferior Negro race."

You would think that some thought might have been given to asking why Black Africans scored lower on Galton's intellectual assessments than White Europeans did, and that someone would suggest such obvious answers as opportunity for education, cultural biases in the assessment tool, and socioeconomic level.  Surprisingly, few did.  The outcome for the Western European elites -- "we're inherently better than the people we're colonizing" -- was so convenient to their goals that it was easier not even to ask the question.

Of course, it bears mention that Galton didn't just hate Black Africans.  He kind of hated everyone who wasn't a member of the English aristocracy.  One of his more astonishing "studies" was a "beauty map" of the United Kingdom, which purported to measure the average beauty of women across the UK, ranking places from the most beautiful to the ugliest.  (The low point, if you're curious, was Aberdeen, Scotland.  Being partly of Scottish descent, I'd like to send a personal memo to Galton to kiss my Celtic ass.)

In 1904, Galton founded the Eugenics Record Office, and along with another person of similar mindset -- his student Karl Pearson -- launched a journal called the Annals of Eugenics (which is still around, but has been rebranded as the Annals of Human Genetics).  Pearson made a huge contribution to the statistical study of genetics, developing methods still in use today.  But he was also responsible for scary stuff like this:

History shows me one way, and one way only, in which a high state of civilization has been produced, namely, the struggle of race with race, and the survival of the physically and mentally fitter race.  If you want to know whether the lower races of man can evolve a higher type, I fear the only course is to leave them to fight it out among themselves, and even then the struggle for existence between individual and individual, between tribe and tribe, may not be supported by that physical selection due to a particular climate on which probably so much of the Aryan's success depended...  No degenerate and feeble stock will ever be converted into healthy and sound stock by the accumulated effects of education, good laws, and sanitary surroundings.  Such means may render the individual members of a stock passable if not strong members of society, but the same process will have to be gone through again and again with their offspring, and this in ever-widening circles, if the stock, owing to the conditions in which society has placed it, is able to increase its numbers.

I'd like to be able to give you the comforting message that the racism, bigotry, and flawed use of statistics Galton and Pearson excelled at have disappeared, but it's still with us.  The 1994 book by Richard Herrnstein and Charles Murray, The Bell Curve, was little more than a modern reworking of Galton and Pearson.  Despite it receiving enormous amounts of criticism from researchers in cognitive psychology, it's widely credited with influencing our current generation of white supremacists, such as Steve Bannon, Stephen Miller, and Elon Musk.

If you still don't believe me, consider a story that just broke last week, which is the reason the whole topic comes up -- that National Institute of Health genetic data on twenty thousand children have been given to a "group of fringe researchers" who turned around and used it to produce sixteen spurious papers claiming to show a genetic (and racial) basis for intelligence.  It's a breach of both privacy and scientific ethics -- not that this is uncommon given the current regime here in the United States -- and shows that although Francis Galton died over a hundred years ago, his twisted spirit lives on.

Even Quetelet, though, should raise some eyebrows.  What, exactly, does it mean to be average?  I remember having that discussion with my principal during my teaching years.  Suppose a particular kid gets a 75% on a test, and that's the average for the class.  I've seen kids score like that when they were very good at regurgitation of facts (so they got all the questions requiring rote memory correct, but few of the deeper ones) and conversely, from kids who were great at understanding the bigger picture in depth, but had issues with recalling terminology.  How can we justifiably throw those two, very different, groups of students into the same bin, stamped with the same all-important number?

As someone on the neurodivergent end of things, I can vouch for the fact that grades don't really mean much.  I'm definitely not Quetelet's homme moyen, and kind of never have been.  I've got a decent brain, but my grades -- especially in high school and the first two years of college -- weren't all that great.  There were a lot of reasons for that -- perhaps a story for another time -- but my point here is the numbers supposedly characterizing me didn't, perhaps, say everything there was to be said about me intellectually.

Our desire to turn everything into numbers has a long and sketchy history, because so few people stop and ask why the numbers are what they are.  Quetelet's legacy misleads us most, I think, in believing that reality can be captured in data alone.  The world is a complex place, and converting it into a handful of statistics may make it seem simpler. 

But at the same time, it also falls far short.  As Ursula LeGuin put it, "I never knew anyone who found life simple.  I think a life or a time looks simple only if you leave out the details."

****************************************