The
lucky cat, a traditional Chinese symbol of good fortune. Photo by Juliane
Zielonka.
In this obscure treatise on abstract mathematics (Aproximatio), written
in Latin nearly two centuries ago, and supposed by its author to have no
practical implications outside the realm of games of chance, in this brief
supplement now so rare that only two copies have been reported, we have
the first formulation of the momentous concept of the law of errors. We
like to imagine De Moivre sitting at a table with a small group of cronies
in some London coffee house, imparting to them the discovery which he had
made through his knowledge of the properties of infinite series and his
interest in the fortunes of the gambler. And we like to imagine the astonished
incredulity with which they would have greeted a prophecy that there would
come a time with when that theorem would powerfully affect the thinking
of the world on all its social problems, when it would enter the schools
and shape the policies of educators, and when its aid would be invoked
in thousands of investigations in sciences whose very name where then unknown.
(p. 14)
[click here for Abraham De Moivre's article, "On the Law of Normal Probability"]
Educational statistics is the offspring of a varied ancestry. The greed
of ancient kings enumerating their people for taxation; the panic of an
English sovereign during the London plague; the cupidity of professional
gamblers; the scientific ardor of psycho-physicists; the labors of mathematicians
and astronomers and physicists and actuaries; the enthusiasm of the students
of social phenomena; the disciplined imagination of the biologists; and
the vision of educators planning a new science of education; from these
has educational statistics descended.
The rapidity with which statistics is becoming
the language of educational research makes it appear highly desirable that
there should be available to students of statistics some account of the
historical development of topics closely related to educational statistical
theory and practice. At present there is but little accessible material
except the original documents from which a student from which may learn
by whom any particular tremor formula or technique was first used, and
what its subsequent history has been. [This remained the case until
the 80s. The subject is still not taught historically!] There is a
prevailing idea that much of statistical theory is of more recent origin
than is actually the case. Formulas published a hundred years ago are spoken
of as though they were the discoveries of this century, if not of this
decade. The same formulas are often discovered independently by different
men who are each unaware of the work of the other. Unnecessary and redundant
terms have been coined, and terms have been used in a sense never intended
by their originators, largely because the field is wide, some of the sources
are difficult to obtain, and not many people have been to take the time
to ascertain exactly what their predecessors have written.
This study attempts to present the modern use of
statistics against a background of the work of De Moivre, Bernoulli, Gauss,
Laplace, Quetelet, Galton, Ebbinghaus, Fechner, and many others. It is
hoped that such a survey may be of use to students in various ways. It
will furnish a historical perspective which will enable them to better
to understand and evaluate the present statistical practices and techniques
employed in professional education, and it will provide a list of sources
helpful to any student desiring to do intensive reading on a particular
topic. It may also be of service to the man about to publish a new formula
or a new term, by helping him discover what terms and formulas have already
been proposed.
… The use of mathematical symbols and formulas has been avoided wherever
feasible, but it is manifestly impossible to discuss the history of statistical
theory in language that is entirely non-technical. (pp. 1-2)
In speaking of Florence Nightingale, Pearson says: ‘Her statistics were more than a study, they were indeed her religion. For her, Quetelet was the hero scientist, and the presentation copy of his Physique Sociale is annotated by her on every page. Florence Nightingale believed – and in all the actions of her life acted upon that belief – that the administrator could only be successful if he were guided by statistical knowledge. The legislator – to say nothing of the political – too often failed for want of this knowledge. Nay, she went further: she held that the universe – including human communities – was evolved in accordance with a divine plan; that it was man’s business to endeavor to understand such a plan and guide his actions in sympathy with it. But to understand God’s thoughts, she held we must study statistics, for these are the measure of his purpose. Thus the study of statistics was for her a religious duty. (p. 43)
Walker, Helen M., Studies in the History of Statistical Method: With Special Reference to Certain Educational Problems, The Williams & Wilkins Company, Baltimore, 1929.
