Who Is Sir Francis Galton?

Sir Francis Galton has given us a living legacy unlike any other. This is simply because his ideas and in particular his gift of “statistical thinking” are nearly universal in their applicability to ongoing enterprises.

In October of 1989, I had travelled to London to see the original Quincunx. And in September of 1991, I returned to London for the Annual Symposium of the Galton Institute. Two days devoted to:

SIR FRANCIS GALTON, FRS (1822 - 1911)

- The Legacy of His Ideas -

Let me quote from the Institute Newsletter announcement of the Symposium: “As Galton had one of the widest ranging minds of the 19th century and studied many topics whose importance has grown in the light of subsequent developments, this symposium promises to be of particularly wide interest.”.

The symposium was a great event for me. I was able to meet the president of the Institute, Professor Margaret Sutherland, and the general secretary, Mrs Linda Brooks, who seemed to know that a lone American could use a little extra attention.

I met a Darwin, (of course, a relative of Sir Francis Galton), and sat next to Sir Andrew Huxley, who was very gracious. He was then 74, and the grandson of Thomas Henry Huxley, the great Victorian biologist and advocate of evolution. T H Huxley was a close associate of Francis Galton throughout their lives.

One of the highlights was a presentation by Dorothy Middleton, a grand niece of Sir Francis Galton, who remembered holding Sir Francis’ hand when she was very young! Not the least diminished by her age, in the eighties, she gave a spirited review of Galton’s travels, and his contributions to Geography, the subjects which brought him notice for the first time in scientific circles.

I got to shake Ms Middleton’s hand while I congratulated her on her presentation. Imagine! Holding the hand that had held the hand of Francis Galton!

And my friend June Rathbone brought over the rather large collection of Galtoniana, and put it on display. Then I had the privilege of having lunch with June. As we say in Texas, “It doesn’t get any better than this.”

The proceedings of the symposium were published in 1993, with the same title as the symposium: “Sir Francis Galton, FRS; the Legacy of His Ideas".

Of the fifteen presentations, four are about Sir Francis’ life from different perspectives.

Ten are about technical subjects; statistics, human growth, twin studies, fingerprints, meteorology, evolution, race, education, intelligence.

One gives the history of the Galton Laboratory, and the last is a reprint of three of Francis Galton’s memoirs. These three had been brought together before in a little moss green pamphlet, published by the Eugenics Society, and the last remaining copies were sold during the seminar. I acquired several at that time, most of which have been given to special friends, or students at my seminars who seemed to be genuinely fascinated by Sir Francis.

The first of these: “Statistical Inquiries into the Efficacy of Prayer” (1872) is still a classic today on how to gather data, how to use it and how to analyse it.

On the dust cover of the published Proceedings is a statement that: “This is .... the definitive book on Francis Galton and his legacy.”. It certainly is a splendid book; a treasure trove for anyone interested in reading and learning about Sir Francis Galton.

However, the Galton Institute is concerned with the study of factors relating to human reproduction, development, and health.

And it will be noted that all of the authors are highly placed members of academia; a more impressive gathering one could hardly imagine.

But the legacy of Sir Francis Galton goes beyond the admirable interests of the Galton Institute, and this part of the legacy is also having a profound effect on human endeavours.

This additional legacy is the use of Sir Francis’ insights in the world of industry, and the yet broader realm of process improvement in all areas of human endeavour. There is at least a small irony here in the fact that Sir Francis’ considerable wealth came from his father, a successful industrialist, whose factory was capable of producing one gun per minute!

A few examples of this “other” legacy follow:

Sir Ronald A Fisher, the second professor of Eugenics and the Galton Professor at UCL, has been said by many to have been the spiritual or conceptual heir to Francis Galton. They both liked to think in terms of distributions, and sought practical uses for their discoveries.

In 1935, Sir Ronald published “The Design of Experiments”, a classic in how to analyse a process, using measurements of more than one variable at a time. These procedures revolutionised experimental methods in agriculture.

The key is to determine whether an observed result is due to intentional change introduced by the experimenter, or is due to random variation; in other words, a result due simply to chance, having nothing to do with the intentional change.

A properly designed experiment reduces the number of trials considerably from what would be required if only one variable at a time is evaluated.

So, for example, for a particular plant, say corn, one can evaluate changes in soil type, seed type, fertiliser type, amount of water, etc, together rather than separately, saving enormous amounts of time and money, and most importantly, arriving at the correct conclusions a large majority of the time.

The following data show the USA agricultural commodity exports and imports for the years 1990 through 1995. For the six years, exports exceeded imports by $113 billion, or an average of $19 billion per year. This is clearly a major plus for the USA economy.

Similar data for cars and trucks shows that the balance of trade for the USA for the years 1989 through 1994 is a negative $263 billion. Exports were less than imports by an average of $44 billion per year.

So the positive trade balance that was achieved in agriculture was less than half the negative trade balance resulting from motor vehicle production.

How could this be? The United States had once been the unchallenged giant in auto production, and in almost every other industrial commodity. How could Japan, a country with a reputation for poor quality, be not only competing, but gaining market share steadily?

Our own Dr W Edwards Deming learned about modern statistics at University College London under Sir Ronald Fisher, the Galton Professor.

At the age of 50, and having a great deal of experience using statistical methods and teaching them, especially during the war effort for WWII, he was summoned to Japan to assist in that devastated country’s recovery.

He was able to convince the leaders of Japanese industry that through the use of statistical thinking, they could win a competition against opponents who used trial and error, and brute force tactics.

This was not accomplished overnight, however. Three decades later, the high quality Japanese car finally made Detroit stop sneering and take notice. And about the same time, some of America’s industry moguls decided it was time to check out this “Dr. Deming”.

There were only a small number of companies, however, that agreed to Dr. Demings’ terms, the main one being the unbendable requirement that the highest level of management must take a leadership role and learn at least the basics of statistical methods.

Ford Motor Co. was one of the first, and their management stuck with the program, and with Dr. Deming. When they began, there were questions about Ford’s survival, but after about ten years of “constancy of purpose”, as Dr. Deming called it, they emerged as the premier force in US auto production, especially in terms of quality and reliability.

The overwhelming conclusion from these striking examples is that the use of statistical thinking imparts a powerful competitive edge.

Now, statistical thinking makes use of many statistical methods. Of these, I want to illustrate the single most powerful method. By this I mean, the method which provides the most information, insight, and guidance for the least amount of effort. This is the control chart, which I will explain fully in my next article.