In article <[EMAIL PROTECTED]>,
Rich Ulrich <[EMAIL PROTECTED]> wrote:
>Here is a bit of a ramble on history and philosophy. It is only
>partly inspired by the single sentence that I am going to
>cite from one post.
>I have been 'doing statistics' in some fashion since 1970, and
>it is only in the last couple of years that I have noticed that
>statistics seems to be a new and developing profession --
>And maybe that is why we don't yet have answers for all the
>questions. (If you have an urge to reply to this but not in public,
>I will welcome e-mail, too.)
>On 2 Jul 2003 16:47:31 GMT, [EMAIL PROTECTED]
>(Radford Neal) wrote: [ I am snipping everything else. ]
>'Meanwhile, statisticians like to think that statistics
>(a) is a mathematically precise discipline, and
>(b) hasn't been barking up the wrong tree for the past 100 years. '
>I'm a statistician. I think that the statistics that I use did not
>exist 100 years ago, for the most part.
>Regression was first described in (IIRC) 1998.
Do you mean 1898? I am not sure what year, but Galton
introduced the term "regression" in the 19th century.
However, using multiple regression (least squares) goes
back to the end of the 18th century, with massive use,
mainly nonlinear, in the 19th. Even the question of
outliers is 19th century. Maximum likelihood goes back
to the middle of the 18th century.
In 1910,
>Pearson was asserting that the 2x2 contingency table
>had 3 d.f., and he was such a browbeating tyrant that it
>was a few years before the 'field' could assert otherwise.
>Fisher wrote his research book in 1928.
Most of what is in that book was done earlier.
Stigler, who writes
>modern books on the history of statistics, has estimated
>that 'modern statistics' got started, more or less, in 1930.
Yes; it was the overthrow of the idea that statistical
significance was related to the truth of the null
hypothesis.
>WW II provided a big stimulus for various sorts of studies
>in statistics and probability. There was code-breaking, and
>how to hunt for submarines (or avoid them).
There was code-breaking, but the new developments in
probability and statistics did not develop enough.
Neither was anything beyond roughly 18th century
probability used in hunting or avoiding submarines.
This may not have been fully used before.
Oh, the whole
>matter of 'national economic statistics' was an invention
>of the war effort, for figuring out how to compare resources.
Try the 17th century for the systematic use of this. The
word "statistics" comes from "state", and gathering state
statistics of this type goes WAY back, such as 11th
century.
>Statistics for quality control were invented during or just
>after the war.
Change that to the 1920s, although much was not published.
However, it was used in industrial quality control in the
1930s, and many of the terms in that go back that far.
Sequential procedures also go back to the 1920s, but the
good ones were invented during the war.
Meehl wrote his book (about 1952) that
>showed that psychiatric/ diagnostic uses of statistics -- or
>of valid numerical logic -- had been, up to that time, almost
>uniformly abominable.
>'Nonparametric statistics' gained over-popularity in the 1950s.
But it goes back 20 years earlier.
>Following disputes about what could or should be said about
>cigarettes causing lung cancer and heart disease, the logic
>of epidemiological evidence was written out in the 1960s.
The first use of statistics for epidemiological evidence
which I know about is the London cholera epidemic of, I
believe, the 17th century.
>Also in the 1960s, the names of Bayes was attached to an
>alternative movement for doing certain things in statistics.
Try the late 18th and early 19th century.
>By the 1970s, computers led to a publication of research
>using multiple variables, along with a swift popularity and
>abuse of 'stepwise' algorithms.
This popularity and abuse goes back at least 40 years
earlier. There was an Iowa State thesis which did a
separate stepwise regression on the effect of weather
on corn yield for each county in the state, going six
steps. Of course, the variables were different in each
of the roughly 90 counties.
A government funded
>initiative had produced the publicly available statistical
>programs in the BMD package, which formed the basis
>(I think) for SPSS and SAS. At this time, just about
>nobody who is practicing statistics or theorizing has had
>a degree in "statistics".
?????
>In 1970, a book was published about the 'controversy
>over statistical tests.' The Forward confessed that the book
>itself was the first organized attempt to stir up controversy,
>since researchers and editors still seemed pretty blithely
>unaware that anyone had any objection or any alternative.
This was an active discussion question 20 years earlier.
For example, look at Blackwell and Girshick, whose book
appeared in 1954, pointing out the problems which had
been discussed in many places starting almost 10 years
earlier.
>(Actually, what still quells the controversy, in my opinion,
>is there's still almost nothing offered as alternative.)
The alternative is there; the problem is that the only
"rational" approach is Bayes, at least prior Bayes.
This follows from simple axioms, and except for one use
of a non-constructive theorem, rather elementary mathematics.
The reasons for its non-acceptance is more theological
than anything else. For one thing, it is not "objective",
and furthermore, it CANNOT be objective.
................
--
This address is for information only. I do not claim that these views
are those of the Statistics Department or of Purdue University.
Herman Rubin, Deptartment of Statistics, Purdue University
[EMAIL PROTECTED] Phone: (765)494-6054 FAX: (765)494-0558
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
